JP5152570B2 - Automotive user hospitality system - Google Patents

Description

  The present invention is intended to assist the user in using the car or to entertain (or serve) the user at least one of approaching, riding, driving, getting off and leaving the car of the user. Relates to the hospitality system for automobile users.

JP 2001-252265 A JP 2007-145225 A JP 2007-145226 A

  In Patent Document 1, a user's current mental state is detected, and when the user himself / herself inputs a target mental state via a man-machine interface, a stimulus is applied to change to the target mental state. A technique for performing feedback control according to the monitoring result is disclosed.

  However, in the technique of Patent Document 1, since the user himself / herself specifies the target mental state, that is, the direction of stimulus application, the device side must respond if the intention is not clearly displayed after the input. Since there are complaints about not being able to get (that is, if you don't ask for something), the stimulus content coming from now on will also be within the prediction range, so there is a difficulty in getting the pleasure of being hospitalized. In addition, since only the target mental state input by the user is prioritized, there is a problem that the mental induction effect according to the current state of the user cannot be achieved appropriately.

  The problem of the present invention is that the direction of mental guidance according to the current mental state is determined spontaneously on the basis of the detected current mental state, and the mental induction is performed even if the user does not indicate intention each time. It is an object of the present invention to provide an automobile user hospitality system that can actively receive a series of hospitality operations related to the above from the system side in an appropriate flow according to the situation.

Means and actions / effects for solving the problems

In order to solve the above problems, the automobile user hospitality system of the present invention is:
A plurality of hospitality operation units for performing a hospitality operation to assist a user in using a car or to entertain a user;
User biometric information acquisition means for acquiring user biometric information;
Mental state specifying means for specifying the user's mental state based on the content of the acquired user biometric information,
A hospitality decision-making unit that determines the content of the hospitality operation by the hospitality operation unit based on the specified mental state;
A hospitality execution control unit that controls the operation of the corresponding hospitality operation unit so that the hospitality operation is performed according to the determination content of the hospitality decision making unit,
The hospitality decision-making department
A series of pre-determined state transition patterns that guide the user's mental state from the base mental state to the end mental state through one or more progress mental states, each adapted to the transition / guidance between adjacent mental states State transition pattern storage means for storing a plurality of patterns in association with the content of the guidance hospitality operation;
A state transition pattern selecting unit that selects a state transition pattern having a base mental state that matches the current mental state of the user specified by the mental state specifying unit, with the state transition pattern storage unit;
Guided hospitality operation command means for sequentially reading out the contents of guided hospitality operation according to the selected state transition pattern, and causing the hospitality execution control unit to sequentially perform the guided hospitality operation to the corresponding hospitality operation unit. and, with a,
The mental state identifying means includes a predetermined neutral state, an anger / excited state in which the mental activity is higher in an uncomfortable direction than the neutral state, and a mental activity level in the unpleasant direction than the neutral state. And discriminating from the low discouragement state,
The state transition pattern storage means, as a state transition pattern with the anger / excited state or the discouraged state as the basic mental state, after passing through the intermediate calmed state corrected to the unpleasant side basic mental state in the pleasant direction, Memorize the modified deliberate state transition pattern leading to the neutral state,
The modified dedicated state transition pattern has a first-guided hospitality movement that shifts from the uncomfortable base mental state to the intermediate calmed state. If the base mental state is angry / excited, it is calmed down, and the basic mental state If the person is discouraged, a counter-type hospitality action that excites this is incorporated into the first induction hospitality action,
The modified dedicated state transition pattern has a preparatory hospitality operation that precedes the first guidance hospitality operation in which a counter-type hospitality operation is incorporated, and a compliant hospitality operation that supports or promotes the base mental state is incorporated in the preparatory hospitality operation. and wherein the Rukoto such Te.

  In the configuration of the present invention, a plurality of state transition patterns that function as scenarios for guiding the mental state of the user are prepared in advance on the hospitality system side (that is, the host vehicle side). Each state transition pattern is determined so as to reach the end point mental state from the base mental state through one or more progress mental states, and a series of guidance hospitality operations respectively adapted to transition / guidance between adjacent mental states. It is stored in association with the contents. Then, the current mental state of the user is detected, and a state transition pattern having a base mental state that matches the detected current mental state is selected, and the contents of the guidance hospitality operation are read and sequentially performed according to the state transition pattern. I made it. As a result, the direction of mental induction is determined spontaneously by the system based on the state transition pattern selected in accordance with the detected current mental state, and as a result, even if the user does not display intentions one by one, The hospitality operation can be actively received from the system side in an appropriate flow according to the situation. Further, by passing through the progress mental state, the user can be comfortably and smoothly led to the end point mental state without feeling unnatural and sudden mental state change.

The mental state identifying means includes a predetermined neutral state, an anger / excited state in which the mental activity is higher in an uncomfortable direction than the neutral state, and a mental activity level in the unpleasant direction than the neutral state. And a low discouraged state can be distinguished from each other. The state transition pattern storage means is a state transition pattern in which the anger / excited state or the discouraged state is a base mental state, and the intermediate calmed mental state obtained by correcting the uncomfortable side base mental state in a pleasant direction is used as a transitive mental state. The modified dedicated state transition pattern leading to the neutral state can be stored. By using the modified dedicated state transition pattern as described above, even if the current mental state is an unpleasant state such as anger / excited state or discouraged state, this is naturally relaxed and it is a stable and calm user mental state. It can lead to a neutral state, and the contents of hospitality will be the basis of a comfortable car life.

  The corrected dedicated state transition pattern can be defined as having a first guiding hospitality operation for transitioning from the unpleasant side base mental state to the intermediate calmed state. This first guidance hospitality action incorporates counter-type hospitality action that calms the base mental state if it is angry / excited and excites it if the base mental state is discouraged. it can. The hospitality movement to calm down in anger / excitement state, and the hospitality movement to excite and raise feelings in the discouraged state, the psychic state on the unpleasant side quickly shifts toward the neutral state Can be made. Such counter-type hospitality operations include scent generation (which is a concept established as aromatherapy) that contributes to calming anger and excitement or activating discouragement, and air conditioning ( If you are in an excited state, lower the air-conditioning temperature, or if you are in a disappointed state, or output sound information (if you are in an irritated state of excitement, sound healing sounds such as healing music and sound effects such as waves) For example, pop music can be output).

  In addition, the modified dedicated state transition pattern can be defined as having a preparatory hospitality operation that precedes the first guidance hospitality operation incorporating the counter-type hospitality operation, and a compliant hospitality operation that supports or promotes the base mental state. Can be incorporated into the preparation hospitality operation. The base mental state is anger / excited state or discouraged state. If so, raise the air conditioning temperature, lower it slightly if you are discouraged, or output sound information (if you are angry and excited, output pop music, heavy metal, hard rock, if you are discouraged, play healing music) And the like. Although this compliant hospitality operation appears to be seemingly contradictory in terms of improving the unpleasant mental state, the system side affirms or justifies the user's unpleasant mental state, so that the user As a preparatory operation that opens and accepts the content of the hospitality operation provided from the above, the effect of resolving the unpleasant mental state of the subsequent counter-type hospitality operation can be further enhanced.

In addition, the first guidance hospitality operation may further include a unique preference suitable type hospitality operation adapted to the user-specific preference regardless of the type of the base mental state. This is a kind of “motivation” for the user by performing a hospitality operation that matches a specific user preference, and is similar to giving a candy or a toy to a crying child. There is an advantage that immediate effect can be expected for improvement of mental state. Such unique taste-preferred hospitality operation can be an output operation of music or visual information that matches the user's preference (for example, user's favorite music or hobby information (image / video) that the user is inclined to) Output operation).

Incorporating both the unique preference-preferred type hospitality operation and the counter-type hospitality operation in the first guidance hospitality operation can resolve the unpleasant mental state more quickly, but from the viewpoint of smoothly transitioning the mental state to the neutral state, As the neutral state approaches, it is desirable to gradually relax mental state corrections in a pleasant direction. In this case, the modified dedicated state transition pattern can be defined as having a second guidance hospitality operation for transitioning from the intermediate calm state to the neutral state, and the second guidance hospitality operation continues from the first guidance hospitality operation. Incorporation of unique- preferred type hospitality operation and the incorporation of the counter-type hospitality operation is more limited than the first guidance hospitality operation (if the counter-type hospitality operation is not incorporated or is incorporated, its activity activity Means lower than the first guidance hospitality operation). In other words, restricting the hospitality operation suitable for the unique taste gives a feeling of loss similar to being able to pick up “confectionery and toys” that the child who once stopped crying takes, and it shifts to the mental state transition to the neutral state On the contrary, it may be counterproductive. Therefore, as described above, the unique preference suitable type hospitality operation is continued, and the counter type hospitality operation is suppressed, so that a smooth transition to the neutral state can be achieved.

Also, the second guidance hospitality operation is based on the premise that multiple users are present, and the incorporation of the counter-type hospitality operation for specific users is more limited than the first guidance hospitality operation, and is common to those multiple users. The common preference suitable type hospitality operation that suits the preference may be incorporated in a form that is promoted more than the first guidance hospitality operation. The purpose of this mode is to determine how smoothly the mental state of a specific user among them is finally transferred to the neutral state. When the neutral state is approaching, there is a possibility that when becoming completely restrict specific preference adaptive hospitality operation user sense of loss becomes large, but as previously described, the inherent taste adaptive hospitality The behavior is set to a common taste-preferred type hospitality operation that can be enjoyed in common by a plurality of users including the specific user, and on the other hand, by suppressing the counter type hospitality operation, the mental state of the specific user can be It is softened in a sense of solidarity with the user, and a smooth transition to the neutral state can be achieved.

On the other hand, in the first guidance hospitality operation, it is effective to perform intensive “motivation” by the unique preference suitable type hospitality operation unique to the user in anticipation of immediate effect. In this case, the second guidance hospitality operation can achieve a smooth transition to the neutral state by limiting the incorporation of the unique preference suitable type hospitality operation more than the first guidance hospitality operation.

  The mental state specifying means includes a user's mental state, a predetermined neutral state, a rising state having a higher mental activity than the neutral state, and a mental activity having a higher direction than the neutral state. A low healing state can be distinguished from each other. The state transition pattern storage means can store an improved dedicated state transition pattern that leads to a swell state or a healed state with the neutral state as a base mental state. Then, the state transition pattern selection means can select the enhanced dedicated state transition pattern subsequent to the selection of the modified dedicated state transition pattern. After transitioning the unpleasant mental state to the neutral state, and further leading to a climax state in which the mental activity is higher in a pleasant direction than the neutral state, it can lead the user who is in a bad mood to a bright and cheerful state, You can make your stay in the car more enjoyable. On the other hand, if you lead to a healed state that has lower mental activity than the neutral state, you can smoothly convert the mental state of a user who is tired from discouragement and anger to a mental state full of openness by healing and stay in the car Can be produced more comfortably.

  In the system of the present invention, when a plurality of users are present, the current mental state is specified for each user, the corresponding state transition pattern is selected, and the guidance hospitality operation is individually performed for each user according to each state transition pattern. Can be configured. There is a tendency that passengers who are not involved in driving can enjoy the effects without hesitation in the hospitality operation that leads to a high mental state in the above-mentioned pleasant direction. In this case, the process of selecting the improved dedicated state transition pattern following the selection of the modified dedicated state transition pattern may be performed only for the user of the seat excluding the driver's seat. Conversely, the driver can ensure driving safety by maintaining a stable mental state, so once the unpleasant mental state is changed to the neutral state, the hospitality operation to maintain the neutral state as much as possible It can be said that it is desirable.

  However, if the neutral state lasts too long, there is a concern that the tension will fade and the attention may be distracted. Therefore, the state transition pattern storage means, as a state transition pattern with the neutral state as the base mental state, returns to the end state mental state with the neutral state through the progress mental state with the swelled state or the healed state. It can be formed to store patterns. According to such a regressive state transition pattern, the mental state can be activated by actively stimulating the mental state of the user (especially the driver) who is in the neutral state, and the state of tension can be activated. Can be achieved. And if a certain sense of tension can be recovered, the tension can be relieved by returning to the neutral state again. On the other hand, if the driving time is prolonged and there is a concern about fatigue, the driver is relaxed by shifting to a healed state. To relieve fatigue.

  The guidance hospitality operation command means can be configured to perform a guidance hospitality operation command for causing the hospitality execution control unit to periodically and repeatedly perform the guidance hospitality operation contents according to the regression type state transition pattern. In other words, by repeating the transition of hospitality movement “neutral state → swelling state (tension / wakefulness) → neutral state → healed state (relaxation / fatigue relaxation) → neutral state”, long tension and relaxation are cycled. By controlling the mental state through the entertaining hospitality action, it is possible to effectively prevent the mental state of the driver from becoming ruined even during driving over time, and greatly contribute to safe driving. In addition, even if the mental stimulation by leading to a state of excitement lasts for an excessively long time, the desired effect may not be improved due to a decrease in activity due to “habituation”. In consideration of this point, the regression type state transition pattern may be set so that one cycle when the progress mental state is healed and set to a state is longer than one cycle when the progress mental state is set to a rising state. This is desirable.

  Moreover, in the regression type state transition pattern, more specifically, as a guidance hospitality operation for transitioning from the neutral state to the healed state, the scent generation operation and the air conditioning operation that contribute to improving the user's self-concentration state Incorporating self-concentration promotion hospitality operations consisting of at least one of lighting output operation and sound information output makes it smoother to the final healing state (relaxed state) once through the self-concentrating state Can lead to. On the other hand, the regression type state transition pattern is a guidance service for transitioning from the neutral state to the excitement state. It is good to incorporate. The common taste-preferred type of hospitality operation can increase the sense of solidarity with the passengers through preference, and as a result, has the advantage of being more easily guided to a rising state.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings. FIG. 1 is a conceptual block diagram of an automobile user hospitality system (hereinafter also simply referred to as “system”) 100 showing an embodiment of the present invention. The system 100 includes a hospitality execution control unit 3 including a first computer to which various hospitality operation units 502 to 517, 534, 541, 548, 549, 550, 551, 552, and 1001B are connected, and various sensors and The vehicle-side mounting unit 100 including the hospitality decision-making unit 2 including the second computer to which the camera groups 518 to 528 are connected is configured as a main part thereof. Each of the first computer and the second computer includes a CPU, a ROM, and a RAM, and various functions described later are realized by executing control software stored in the ROM using the RAM as a work memory.

  In the system 100, a series of operations related to the user's use of the car until the user approaches the car, gets into the car, drives the car or stays in the car, and then gets off the vehicle. The scene is divided into a plurality of predetermined scenes. Then, for each of a plurality of divided scenes, the hospitality operation units 502 to 517, 534, 541, 548, 549, 550, 551, 552, and 1001B assist the user in using the car or enjoy the user. The hospitality operation for the purpose is performed. In the present embodiment, a horn 502 and a buzzer 503 are connected as a sound wave generator to the outside of the vehicle. Further, as a lighting device (lamps), a head lamp 504 (beam can be switched between high and low), fog lamp 505, hazard lamp 506, tail lamp 507, cornering lamp 508, backup lamp 509, stop lamp 510, interior lighting 511 and an underfloor lamp 512 are connected. In addition, as other hospitality operation units, an air conditioner 514, a car audio system (car stereo) 515, an electric seat / handle 516, an angle adjustment drive unit 517 such as a side mirror and a rearview mirror, a car navigation system 534, and a door opening / closing operation Assist mechanism (hereinafter referred to as a door assist mechanism) 541, a fragrance generating section 548 that releases a fragrance into the vehicle, and a generating section 549 for generating a consciousness / awakening component for a severe physical condition (including a state in which severe drowsiness is caused) For example, it is attached to the center of the driver's handle so as to spout a component for awareness and awakening (for example, one containing ammonia) in the form of aiming near the driver's face), alerting the driver, Seat vibrator 550 (was embedded in the seat bottom or backrest) for awakening from drowsiness ) Is attached to the shaft of the handle vibrator 551 (handle), the noise canceller 1001B for reducing cabin noise is connected.

  FIG. 2B shows a configuration example of the interior illumination 511, and a plurality of illumination units each having a unique illumination color (in this embodiment, red illumination 511r, amber illumination 511u, yellow illumination 511y, white illumination 511w). And blue-based illumination 511b). These lights are selected by receiving a control command signal input from the hospitality decision making unit 2 through the hospitality execution control unit 3, and specified lighting is selected, and lighting control is performed in various lighting patterns according to the control command signal. . FIG. 3 shows a configuration example of the lighting control data 402 determined according to the personality type of the user. The lighting control data 402 is stored in the ROM of the hospitality decision making unit 2 and is read and used as needed by the control software. For example, for the active personality (SKC1 (see FIG. 11)), the red illumination 511r is selected and flashed (first only, then continuous lighting), and for the gentle personality (SKC2), For example, the amber system lighting 511u is selected and fade-in lighting is performed, but this is only an example.

  In addition to the incandescent light bulb and the fluorescent lamp, the lighting device using a light emitting diode can be adopted as the lighting device. In particular, various illumination lights can be easily obtained by combining light emitting diodes of three primary colors of red (R), green (G), and blue (B). FIG. 4 shows an example of the circuit configuration. Each of the red (R), green (G), and blue (B) light emitting diodes 3401 is connected to a power source (Vs) and is switched by a transistor 3402. Driven. This switching is PWM controlled by a duty ratio determined by the period of a triangular wave (which may be a sawtooth wave) input to the comparator 3403 and the voltage level of the command signal. The input waveform of the command signal to the light emitting diode 3401 of each color can be changed independently, and an illumination color of an arbitrary color tone can be obtained according to the mixing ratio of the three emission colors, and the color tone and illumination intensity pattern can be changed. It is also possible to change over time according to the input waveform of the command signal. Note that the light emission intensity of each color light emitting diode 3401 can be adjusted at the drive current level on the premise of continuous lighting, in addition to the PWM control method as described above, and this is combined with PWM control. A scheme is also possible.

  FIG. 5 shows the relationship between the mixing ratio (depending on the duty ratio) of red (R), green (G), and blue (B) light and the color of the mixed light that is visually recognized (here, The mixing ratio shown represents “1” as a relative mixing ratio of other colors to the set color, and the absolute illumination intensity is set separately based on this relative mixing ratio). Each is assigned an index (0 to 14) for selecting the emission color at the time of control, and serves as control reference information in the ROM of the hospitality execution control unit 3 (or the storage device 535 on the hospitality decision making unit 2 side: necessary for control) Information may be transmitted to the hospitality execution control unit 3 by communication). The white illumination light is frequently used, and the index that defines the white color is determined so as to appear periodically on the array of the index so that the control can be smoothly transferred to and from the colored illumination light. Yes.

  In particular, with the white (index: 6) located in the middle as the boundary, warm colors (light yellow → yellow → red) and cold colors (light blue → blue → blue purple) are arranged before and after that, and the physical condition of the user Depending on the condition and mental state, it is possible to smoothly switch from white illumination light to warm color illumination light or cold color illumination light. Here, the illumination color at the normal time not particularly considering the production or the like is set centering on white. And the white color is selected in the mental state of the middle, and it is determined that the illumination color changes to the blue system, that is, to the short wavelength side, the red color, that is, to the long wavelength side, as the mental state that sinks, to the blue system, the short wavelength side. ing.

  Next, FIG. 6 shows a configuration example of the car audio system 515. The hospitality music performance control information such as music identification information and volume control information is sent from the hospitality decision making unit 2 to the hospitality execution control unit 3. An input interface unit 515a is included. The interface unit 515a is connected to a digital audio control unit 515e and music source databases 515b and 515c (the former is an MP3 database and the latter is a MIDI database) storing a large number of music source data. The music source data selected based on the song identification information is sent to the audio control unit through the interface unit 515a, where it is decoded into digital music waveform data, converted into analog by the analog conversion unit 515f, and then pre-amplifier 515g and power amplifier. After 515h, the sound is output from the speaker 515j at the volume specified by the hospitality music performance control information.

  Returning to FIG. 1, the electric door mechanism 541 is for automatically opening / closing or opening / closing power assisting a sliding door for swinging or a swing type door by a motor (actuator) (not shown).

  FIG. 7 is a functional block diagram illustrating a configuration example of the noise canceller 1001B. The main part of the noise canceller 1001B includes an active noise control mechanism main body 2010 that forms noise suppression means, and a necessary sound enhancement part (means) 2050. The active noise control mechanism 2010 is a noise that synthesizes an in-vehicle noise detection microphone (noise detection microphone) 2011 that detects noise invading into the vehicle, and a noise control waveform that is opposite in phase to the noise waveform detected by the in-vehicle noise detection microphone 2011. And a control waveform synthesis unit (control sound generation unit) 2015. The noise control waveform is output from the noise control speaker 2018. There are also provided an error detection microphone 2012 for detecting an unerased noise component included in the vehicle interior sound after the noise control sound wave is superimposed, and an adaptive filter 2014 for adjusting a filter coefficient in a direction in which the level of the unerased noise is reduced. ing.

  Vehicle interior noise having a sound source in the vehicle itself includes engine sound, road surface sound, wind noise, etc. A plurality of vehicle interior noise detection microphones 2011 are distributed and arranged at positions suitable for detection of individual vehicle interior noise. ing. The in-vehicle noise detection microphones 2011 are at different positions as viewed from the passenger J, and there is a considerable phase difference between the noise waveform at the position picked up by the microphone 2011 and the noise waveform actually heard by the passenger J. . Therefore, in order to match this phase difference, the detection waveform of the vehicle interior noise detection microphone 2011 is appropriately supplied to the control sound generation unit 2015 via the phase adjustment unit 2013.

  Next, the necessary sound enhancement unit 2050 includes an enhanced sound detection microphone 2051 and a necessary sound extraction filter 2053, and the extracted waveform of the necessary sound is given to the control sound generation unit 2015. Also here, the phase adjustment unit 2052 is provided as appropriate due to the same situation as the vehicle interior noise detection microphone 2011. The emphasis sound detection microphone 2051 includes an in-vehicle microphone 2051 for capturing necessary sounds outside the vehicle and an in-vehicle microphone 2051 for capturing necessary sounds in the vehicle. Both can be configured with known directional microphones, and are mounted on the outside of the vehicle so that the angle range with strong directivity for sound detection faces the outside direction of the vehicle and the angle range with low directivity faces the inside direction of the vehicle. In the present embodiment, the entire microphone 2051 is attached so as to go out of the vehicle, but the inside area of the vehicle is set so that the angle area with low directivity is located inside the vehicle and only the angle area with strong directivity goes out of the car. It can also be installed across the outside of the vehicle. On the other hand, the in-vehicle microphone 2051 has a low directivity because the sound detection directivity angle range faces the front of the passenger so that the passenger's conversation sound can be selectively detected corresponding to each seat. It is attached so that the angle range faces in the opposite direction. Each of these emphasized sound detection microphones 2051 is connected to a necessary sound extraction filter 2053 that preferentially passes a necessary sound component in its input waveform (detected waveform). Note that the audio input of the car audio system 515 in FIG. 6 is used as the in-vehicle sound source 2019. The speaker output sound of this audio device (the speaker may be used also as the noise control speaker 2018 or may be provided separately) is controlled so as not to be canceled even if the noise control waveform is superimposed.

  FIG. 8 shows an example of a hardware block diagram corresponding to the functional block diagram of FIG. A first DSP (Digital Signal Processor) 2100 constitutes a noise control waveform synthesis unit (control sound generation unit) 2015 and an adaptive filter 2014 (and further a phase adjustment unit 2013), and an in-vehicle noise detection microphone 2011 is a microphone amplifier. 2101 and an A / D converter 2102, and a noise control speaker 2018 are connected via a D / A converter 2103 and an amplifier 2104, respectively. On the other hand, the second DSP 2200 constitutes an extraction unit for the noise component to be suppressed, and the error detection microphone 2012 is not subject to suppression through the microphone amplifier 2101 and the A / D converter 2102 and is not subject to suppression such as audio input. An audio signal source, that is, a necessary sound source 2019 is connected via an A / D converter 2102.

  The necessary sound enhancement unit 2050 includes a third DSP 2300 that functions as a necessary sound extraction filter 2053, and a necessary sound detection microphone (emphasis sound detection microphone) 2051 is connected via a microphone amplifier 2101 and an A / D converter 2102. Yes. The third DSP 2300 functions as a digital adaptive filter. The filter coefficient setting process will be described below.

  Caution or danger should be recognized for emergency vehicles (ambulances, fire engines, police cars, etc.) sirens, railroad crossing alarms, horns of subsequent vehicles, whistle sounds, and human screams (such as children's screams and women's screams). It is determined as a necessary vehicle exterior sound (emphasized sound), these sample sounds are recorded on a disk or the like, and are made into a library as reference enhanced sound data that can be read and reproduced. As for the conversation sound, a plurality of individual model sounds are similarly stored in a library as reference emphasized sound data. In addition, when the boarding candidate is fixedly fixed, if the model voice is prepared as reference enhancement sound data by the utterance of the model voice itself, the boarding candidate will get on Can enhance the accuracy of conversation sound.

  Then, an appropriate initial value is given to the filter coefficient, and the enhancement sound detection level by the enhancement sound detection microphone 2051 is set to the initial value. Next, each reference emphasis sound is read and output, and is detected by the emphasis sound detection microphone 2051. Then, the passing waveform of the adaptive filter is read, and the level of the waveform that has passed as the reference enhancement sound is measured. The above process is repeated until the detection level reaches the target value. In this way, the filter coefficient is subjected to learning processing so that the reference emphasis sound is successively replaced for both the vehicle exterior sound and the vehicle interior sound (conversation sound), and the detection level of the passing waveform is optimized. The necessary sound is extracted from the input waveform from the emphasized sound detection microphone 2051 by the necessary sound extraction filter 2053 having the filter coefficient adjusted as described above, and the extracted enhanced sound waveform is transferred to the second DSP 2200. The second DSP 2200 calculates a difference between the input waveform from the necessary sound source (audio output here) 2019 and the extracted enhanced sound waveform from the third DSP 2300 from the detection waveform of the vehicle interior noise detection microphone 2011.

  The filter coefficients of the digital adaptive filter incorporated in the first DSP 2100 are initialized prior to use of the system. First, various noises to be suppressed are determined, and those sample sounds are recorded on a disk or the like, and are made into a library as reproducible reference noise. Then, an appropriate initial value is given to the filter coefficient, and the noise level left by the error detection microphone 2012 is set to the initial value. Next, the reference noise is sequentially read out and output, and detected by the vehicle interior noise detection microphone 2011. The detection waveform of the vehicle interior noise detection microphone 2011 that has passed through the adaptive filter is read and subjected to fast Fourier transform, thereby decomposing the noise detection waveform into sine wave elementary waves having different wavelengths. Then, an inverted elementary wave obtained by inverting the phase of each sine wave elementary wave is generated and synthesized again, thereby obtaining a noise control waveform having a phase opposite to that of the noise detection waveform. This is output from the noise control speaker 2018.

  If the coefficient of the adaptive filter is appropriately determined, only the noise component should be efficiently extracted from the waveform of the vehicle interior noise detection microphone 2011. Therefore, the noise control waveform synthesized based on the noise is used for the interior of the vehicle. Noise can be offset without excess or deficiency. However, if the setting of the filter coefficient is not appropriate, the waveform component that is not canceled out remains as a noise component. This is detected by the error detection microphone 2012. The level of the unerased noise component is compared with the target value. If the level is not lower than the target value, the filter coefficient is updated, and the same processing is repeated until it becomes lower than the target value. In this way, the reference noise is replaced one after another, and the filter coefficient is subjected to learning processing so that the unerased noise component is minimized. During actual use, the noise component that remains unerased is constantly monitored, the filter coefficient is updated in real time so that it is always minimized, and the same processing as described above is performed to leave the necessary sound wave component. Only the noise level in the vehicle can be effectively reduced.

  Returning to FIG. 1, the user terminal device 1 is configured as a known mobile phone in the present embodiment (hereinafter also referred to as “mobile device 1”). The portable device 1 can download ringtone data and music data (MP3 data or MIDI data: also used as a ringtone) for ringtone output and music performance. A performance output based on the data can be performed by the circuit.

The hospitality decision making unit 2 is connected to the following sensor / camera group. Some of these function as scene estimation information acquisition means and also function as user biometric characteristic information acquisition means.
External camera 518: Takes a picture of a user approaching the car. Acquire user gestures and facial expressions as still images or videos. In order to magnify a user for photographing, an optical zoom method using a telephoto lens and a digital zoom method for digitally enlarging a photographed image can be used in combination.
Infrared sensor 519: Takes a thermograph based on infrared rays emitted from the face portion of the user approaching or riding the vehicle. By functioning as a body temperature measurement unit that is a user biometric information acquisition means and measuring the temporal change waveform, it is possible to estimate the user's physical condition or mental condition.

Seating sensor 520: Detects whether the user is seated on the seat. It can be composed of a proximity switch or the like embedded in the automobile seat. In addition, a seating sensor can be configured by a camera that captures a user seated on a seat. With this method, when a load source other than a person such as luggage is placed on the seat, it can be distinguished from a case where a person is seated, for example, a hospitality operation is performed only when a person is seated. Selection control is also possible. In addition, if a camera is used, it is possible to detect the motion of a seated user, and the detection information can be further diversified. In order to detect the user's movement on the seat, there is a method using a pressure sensor mounted on the seat.

  Furthermore, in this embodiment, as shown in FIG. 9, the posture of the seated user (driver) based on the detection outputs of the seating sensors 520A, 520B, and 520C embedded in a plurality of distributed manners in the seat and backrest of the seat. The change is detected as a waveform. Both are configured by a pressure sensor that detects a seating pressure, and specifically, a reference sensor 520A is arranged at the center of the back of a user seated facing the front. The remaining sensor is composed of a left sensor 520B arranged to be deviated to the left side of the seat and a right sensor 520C arranged to be deviated to the right side of the seat. The difference between the output of the reference sensor 520A and the output of the right sensor 520C and the output of the left sensor 520B is calculated by the differential amplifiers 603 and 604, respectively. Input to the amplifier 605. The posture signal output Vout (second-type biological condition parameter) is substantially a reference value (here, zero V) when the user is sitting facing the front, and when the posture is biased to the right, the output of the right sensor 520C is output. Since the output of the left sensor 520C decreases and shifts to the negative side, when the posture is biased to the left, the opposite is true and shifts to the positive side. Note that the right sensor 520C and the left sensor 520B are output by the adders 601 and 602 as an added value of the sensor output on the seat side and the sensor output on the back side, but the remaining sensor output and the back sensor output are output. (When doing so, the output on the backrest sensor side decreases when the driver turns forward, and the difference value increases, so this is detected as a larger posture collapse.) be able to.

-Face camera 521: Photographs the facial expression of the user who is seated. For example, it is attached to a rearview mirror or the like, and the upper body including the face of the user (driver) seated on the seat is photographed obliquely from above from the front glass side. Various facial expressions shown in FIG. 10 can be specified by cutting out an image of the face portion from the image and comparing it with a master image prepared by photographing various facial expressions of the user in advance. In either mental state or physical condition, the order of facial expressions is determined in the order of good condition, and points are assigned according to the order (for example, in the case of mental state, stability is “1”, distraction and anxiety “2”, excitement / anger is set to “3”, etc.), facial expressions can be used as discrete numerical parameters (second-type biological state parameters), and their temporal changes can be measured as discrete waveforms. Based on the waveform, it is possible to estimate a mental state or a physical condition. Note that a change in the posture of the driver can also be detected from the image shape of the upper body including the face and the position of the center of gravity on the image. That is, the change waveform of the center of gravity position can be used as a change waveform of posture (second-type biological condition parameter), and it is possible to estimate a mental state or a physical condition based on the waveform. In addition to the function as an acquisition source (user biometric characteristic information acquisition means) of user biological state information used for hospitality control, it is also used for user authentication by biometrics. It is also possible to specify the direction of the face and line of sight by detecting the direction of the iris of the eyes (for example, if you look at the direction of the clock at times, it is estimated that you are impatient with time) Etc.) Further, based on the temporal change waveform of the angle of the line-of-sight direction (detecting the shake angle to the left and right with respect to the reference direction as a reference direction when the head is facing directly in front as the waveform change) (second-type biological condition parameter), It is also used to estimate the physical condition or mental state of the driver.
Microphone 522: Detects user voice. This can also function as user biometric characteristic information acquisition means.

Pressure sensor 523: It is attached to the grasping position of the steering wheel or shift lever of the automobile by the user, and detects the gripping force of the user, the repetition frequency of gripping and releasing, etc. (user biometric information acquisition means).
Blood pressure sensor 524: Attached to the user grasping position of the steering wheel of the automobile (user biological characteristic information acquisition means). The blood pressure value detected by the blood pressure sensor 524 is recorded as a waveform of the temporal change (first-type biological condition parameter), and is used for estimating the physical condition or mental condition of the driver based on the waveform.
Body temperature sensor 525: a temperature sensor attached to a user grasping position on the steering wheel of the automobile (user biometric information acquisition means). The body temperature value detected by the body temperature sensor 525 is recorded as a waveform of the temporal change (first type biological condition parameter), and is used to estimate the physical condition or mental state of the driver based on the waveform.
Skin resistance sensor 545: a well-known sensor that measures the resistance value of the body surface due to perspiration or the like, and is attached to the user grasping position of the steering wheel of the automobile. The skin resistance value detected by the skin resistance sensor 545 is recorded as a waveform of the temporal change (first type biological condition parameter), and is used to estimate the physical condition or mental condition of the driver based on the waveform.

Retina camera 526: Takes a user's retina pattern and is used for user authentication by biometrics.
Iris camera 527: A camera is attached to a rearview mirror or the like, takes an image of the user's iris (iris), and is used for user authentication by biometrics. When an iris image is used, collation / authentication is performed using the personality of the pattern and color. In particular, the iris pattern is an acquired component and has a low genetic influence, so there is a significant difference even in identical twins, and there is an advantage that it can be reliably identified. An authentication method using an iris pattern has features that it can quickly recognize and collate and has a low misidentification rate. Further, the physical condition or the mental condition can be estimated based on the temporal change of the pupil size (second type biological condition parameter) of the driver photographed by the iris camera.
Vein camera 528: The user's vein pattern is photographed and used for user authentication by biometrics.
-Door courtesy switch 537: Detects opening and closing of the door. It is used as a scene estimation information acquisition means for detecting the transition to the boarding scene and the getting-off scene.

  In addition, the output of the ignition switch 538 for detecting engine start is branched and input to the hospitality decision making unit 2. In addition, a humidity sensor 546, a room temperature sensor 563, a sunshine sensor 564 (for controlling the air conditioner 514), a vehicle exterior noise sensor 562 (for controlling the noise canceller 1001B), an illuminance sensor 539 for detecting the brightness level in the vehicle, and an acoustic level in the vehicle. The sound pressure sensor 540 to be measured is similarly connected to the hospitality decision making unit 2.

  Further, the hospitality decision making unit 2 may be a touch panel (a touch panel superimposed on the monitor of the car navigation system 534 may be used: in this case, input information is transferred from the hospitality execution control unit 3 to the hospitality decision making unit 2 And the like, and a storage device 535 configured by a hard disk drive or the like that functions as a hospitality operation information storage unit.

  On the other hand, the hospitality execution control unit 3 is also connected with a GPS 533 (also used in the car navigation system 534), a brake sensor 530, a vehicle speed sensor 531 and an acceleration sensor 532 for acquiring vehicle position information.

  The hospitality decision-making unit 2 acquires user biological state information including at least one of the personality, mental state, and physical condition of the user from detection information of one or more of the sensor / camera groups 518 to 528, and includes the contents thereof In response, it determines what hospitality operation is to be performed by any hospitality operation unit, and commands this to the hospitality execution control unit 3. In response to this, the hospitality execution control unit 3 causes the corresponding hospitality operation units 502 to 517, 534, 541, 548, 549, 550, 551, 552, and 1001B to execute the hospitality operation. That is, the hospitality decision making unit 2 and the hospitality execution control unit 3 cooperate with each other, and the hospitality operation units 502 to 517, 534, 541, 548, 549, 550, The function of changing the operation contents of 551, 552, and 1001B is realized. The hospitality execution control unit 3 is connected to a wireless communication device 4 that constitutes a vehicle side communication means (host side communication means). The wireless communication device 4 communicates with a user-side terminal device (mobile phone) 1 carried by a user of an automobile via a wireless communication network.

  On the other hand, the car audio system 515 is provided with an operation unit 515d (FIG. 6) that is manually operated by a user, and by inputting music selection data from the operation unit 515d, desired music source data can be read and played. The volume / tone control signal from the operation unit 515d is input to the preamplifier 515g. This music selection data is transferred from the interface unit 515a to the hospitality decision making unit 2 via the hospitality execution control unit 3 of FIG. 1, and is stored as music selection result data in the storage device 535 connected thereto. Based on the stored contents, a user-personality determination process, which will be described later, is performed (that is, the operation unit 515d of the car audio system 515 can be said to constitute a function of user biometric characteristic information acquisition means).

  FIG. 11 shows an example of the database structure of the music source data. In the database 401, music source data (MP3 or MIDI) is stored in association with song IDs, song names, and genre codes. In addition, each music source data includes the personality type estimated for the user who selected the music (“active”, “adult”, “optimistic”, “pessimistic”, “degenerate”, “athletic” ”,“ Intellect ”,“ Romanticist ”, etc., as well as the age code (“ Infant ”,“ Children ”,“ Junior ”,“ Youth ”,“ Middle Age ”,“ Midpoint ”,“ Middle Age ”) , “Respect for the elderly”, “age unrelated”, etc.), and gender codes (“male”, “female” and “gender unrelated”) are stored in association with each other. The personality type code is one of the user personality specifying information, and the age code and the sex code are sub-classes unrelated to the personality. Even if the personality of the user can be identified, if the music source that does not match the age group or gender is selected, the effect of “hospitality” to entertain the user is halved. Therefore, in order to further narrow down the suitability of the music source provided to the user, the sub-classification as described above is effective.

  On the other hand, each music source data also stores a song mode code in association with each other. The song mode code is data indicating the mental state and physical condition of the user who selected the song, and the relationship between the song and the song. In the present embodiment, the song mode code is “smooth”, “exhilarating”, “warm / healing”. And “Healing / α-wave system”. Since the personality type code, age code, gender code, genre code, and song mode code are data to be referred to when selecting the hospitality content specific to each user, these are collectively referred to as hospitality reference data. .

  Next, in this embodiment, an approach scene, a boarding scene, a preparation scene, a driving scene, a stay scene, a getting-off scene, a leaving scene, and the like are defined as scenes. Since the method for specifying a scene is disclosed in detail in Patent Documents 1 to 3, it will not be repeated here. A scene can be specified by specifying a relative distance to a user located outside the automobile and its change, and detecting that the user has approached the automobile within a predetermined distance. Further, the boarding scene and the getting-off scene are specified based on the door opening detection output of the door courtesy switch 537. In addition, an individual scene flag corresponding to each scene is provided, and each time a scene whose arrival order is determined in time series arrives, the flag corresponding to that scene is set to “arrival (flag value 1)”. By moving, it is possible to specify which scene is currently advanced. The preparation scene and the scene driving / staying scene are both specified by whether or not the above-described seating sensor detects the user, but until the ignition switch 538 is turned on after entering the automobile, or It is recognized as a preparation scene until the ignition switch 538 is not turned ON and the continuation of sitting beyond a certain level is confirmed. Further, the transition to the leaving scene is identified by the door courtesy switch 537 detecting the door closing after the getting-off scene.

  Each hospitality operation is controlled by the operation control application of the corresponding hospitality operation unit. These operation control applications are stored in the ROM (or storage device 535) of the hospitality execution control unit 3.

  Hereinafter, the operation of the automobile user hospitality system (hereinafter also simply referred to as “system”) 100 will be described. FIG. 12 conceptually shows an overall algorithm of a series of processes from hospitality decision-making to hospitality operation execution in the system 100 (these three figures are connected by using a corresponding circle number as a connector). It should be read as a figure). The hospitality main processes are “objective estimation (δ1)”, “personality adaptation (δ2)”, “state adaptation (δ3)”, “production correspondence (δ4)”, “function selection (δ5)”, “drive (δ6)”. It consists of each step.

  First, in “objective estimation (δ1)”, a current scene is estimated by user position detection (β1) and user motion detection (β2). Specifically, the user position detection (β1) is performed by grasping and specifying the relative positional relationship (α1) between the user and the automobile. Further, the approach direction (α2) of the user can also be considered. On the other hand, the user motion detection (β2) is basically performed by sensors (scene estimation information acquisition means) that detect motions fixed for scene determination, such as door opening and closing operations and seating on a seat. The output is used (α5). Further, the duration (α6) of the specific operation can also be taken into account, as in the case where the transition from the preparation scene to the driving / staying scene is detected based on the seating duration.

  If the scene is determined at γ1, the hospitality intention in each scene is estimated at δ1. As described above, “estimating hospitality intention” is to estimate the hospitality content that best suits the user's physical condition and mental state that changes every time, that is, the type of hospitality content that the user wants most. Specifically, the intention estimation table 371 illustrated in FIG. 19 and FIG. 20 and the principle action table 372 illustrated in FIG. 21 are used in combination, and safety, convenience (easily), and comfort are classified for each scene. For each classification item, a search is made as to whether or not there is a hospitality intention that matches the user's bodily sensation, that is, a control target environment item for each of the tactile system, visual system, olfactory system, and auditory system. If the hospitality intention is retrieved, the hospitality function corresponding to the retrieved hospitality intention is extracted with reference to the corresponding scene-specific function extraction table 373 illustrated in FIG.

  In δ2, the content of the hospitality is adapted to the user's personality. Specifically, this is done by appropriately weighting each individual hospitality process according to the personality determination process described later and the determined personality, that is, in order to adapt to the individual user's personality. The object is to customize the combination of hospitality operations as appropriate, or to change the level of hospitality operations. In order to specify the individuality, personality detection processing β4 is necessary. The personality detection process β4 determines the personality classification more analytically from the method of obtaining personality classification by the user's own input, such as questionnaire processing (α7), and the user's actions, actions, thought patterns or facial expressions. Both methods can be used. About the latter, the system (patent documents 2 and 3) which determines personality classification from the music selection statistics is adopted (see also α8: W2).

  δ3 indicates a process of adapting the hospitality content to the mental / physical state of the user. Further, δ4 is a hospitality effect handling process, and δ5 is a function selection process. That is, mental / physical condition information reflecting the mental state and physical condition of the user is acquired based on the detection information of the user biological characteristic information acquiring means, and the mental state or physical condition of the user is estimated according to the acquired content.

  User biometric characteristic information acquisition means includes an infrared sensor 519 (facial color: α17), a face camera 521 (expression: α9, posture: α11, line of sight: α12, pupil diameter: α13), blood pressure sensor 524 (heartbeat (electrocardiogram): α14. In addition, sensors (502w, 530, 531, 532; erroneous operation rate: α10), blood pressure sensor (α15), seating sensor 520 (pressure sensor) Measure the weight distribution on the seat, detect the minute weight shift while driving, make a judgment that the calmness during driving is impaired, detect how the weight is biased, To determine the degree of fatigue). Details will be described later.

  The output from the above-mentioned user biological characteristic information acquisition means is replaced with a numerical parameter indicating the mental state or physical condition (β5), and the mental state or physical condition of the user is estimated from the numerical parameter and its temporal change (γ3, γ4) ). The estimation result is used for the determination process of the reference intention parameter value described later, and the reference intention parameter value extracts and controls the hospitality function using the intention estimation table 371 (FIGS. 19 and 20) and the principle action table 372 together. Used for content determination processing. The reference intention parameter is set to a specific value that reflects the estimated mental state and physical condition, and the selected function and its control content are adapted to the estimated mental state and physical condition of the user accordingly. Optimized to be a thing. In addition, even if the hospitality of the same scene, if the personality of the user is different, the function is fine-tuned so as to be a hospitality operation adapted to that personality, and even the same user is adopted according to the mental state and physical condition The type and degree of the are adjusted.

  Taking the case of illumination light as an example, the lighting color directed by the user differs depending on the personality (for example, the active type is red, the gentle type is green or blue), and the lighting intensity depends on the physical condition. (For example, when the physical condition is poor, the amount of light is reduced to suppress stimulation by illumination) is often different. The former is hospitality control for adjusting the frequency or wavelength of the illumination light (the wavelength becomes shorter in the order of red, green, and blue), and the latter is hospitality control for adjusting the amplitude of the illumination light. In addition, mental state is a factor related to both of them, and in a somewhat cheerful mental state, it is possible to adopt red illumination light (frequency adjustment) to further boost the mood, and the color of the illumination light It is possible to increase brightness without changing (amplitude adjustment). Also, in the state of excessive excitement, processing such as adopting blue illumination light (frequency adjustment) to reduce feelings or reducing brightness without changing the color of the illumination light (amplitude adjustment) is considered. It is done. In the case of music, since it contains various frequency components, it is more complicated, but in order to enhance the arousal effect, to emphasize the high frequency sound wave of about several hundred Hz to 10 kHz, or conversely sink The control pattern by frequency / amplitude, such as adopting so-called α-wave music that combines the center frequency of sound wave fluctuation with the frequency (7-13 Hz: human resonance) of the brain wave (α wave) when relaxing Can be grasped as well.

  Also, information (disturbance stimulus) on how much stimulus the user currently feels is obtained from the output of the illuminance sensor 539 (visual stimulus: α18), sound pressure sensor (auditory stimulus: α19), etc. in FIG. Environment detection: β6), the disturbance stimulus value is estimated (γ5). In addition, as the disturbance stimulus to be specified, a tactile stimulus (α20: for example, a pressure sensor 523 attached to a handle), an olfactory stimulus (α21: by an olfactory sensor), or the like can be used in combination. Regarding disturbance estimation, indirect stimulation from the space surrounding the user, specifically, height (α22), distance (α23), depth (α24), and the physique of the passenger or passenger (α25), etc. (Spatial detection: β7).

Next, the user personality classification can be determined by the following method, for example.
The user of the car enters in advance a user registration unit 600 (FIG. 2A) formed in a ROM (preferably configured with a flash ROM so that rewriting is possible) of the hospitality decision-making unit 2 in FIG. be registered. This user registration unit is registered in such a manner that each user name (or user ID (and personal identification number) and its personality type are associated with each other, as will be described later. It is also possible to make a decision based on the music selection statistics information of the car audio system that is accumulated during continuous use, but when the music selection statistics information is insufficient, such as immediately after the start of use of the car, or the operation history If you want to estimate the personality type without collecting information, let the user input personality type information or information necessary to identify the personality type information, and determine the personality type based on the input result. You may do it.

  For example, the personality type is displayed on the monitor 536 of FIG. 1 (the monitor of the car navigation system 534 may be substituted), and the user can select the personality type that suits him and input it from the input unit 529. it can. Further, instead of inputting the personality type directly, a method of inputting a questionnaire for determining the personality type may be adopted. In this case, questionnaire questions are displayed on the monitor 536, and the user answers by selecting an answer from the answer options (here, the options are configured with a selection button (not shown), and a touch panel overlaid thereon) Touch the corresponding position of 529 to make a selection input). By answering all the questions, one is uniquely determined from a personality type group determined in advance according to the combination of the answers.

  The user registration input including the user name is also made from the input unit 529, and stored in the user registration unit 600 of FIG. Is done. These series of inputs can also be performed from the portable device 1, and in this case, the input information is transferred to the automobile side by radio. In addition, when the user purchases a car, there is a method in which the dealer side uses the input unit 529 or a dedicated input tool to perform user registration input in advance.

  Since the case of determining the personality type based on the statistical information of the track selection performance of the car audio system is as disclosed in Patent Documents 1 to 3, the outline will be described. In the car audio system 515 shown in FIG. 6, the user can enjoy a performance by selecting a favorite song at any time by an input from the operation unit 515d. When the user selects the music by himself, the specific information (user name or user ID) of the user, the ID of the selected music source data, and the aforementioned hospitality reference data RD (personality type code, age code, gender code, The genre code and the music mode code) are stored in the music selection record storage unit (formed in the storage device 535 in FIG. 1) in a form associated with each other. In the music selection record storage unit, statistical information of the music selection record (stored in the storage device 535 in FIG. 1) is created for each user. In this statistical information, music selection data is counted for each personality type code, and which personality type of music is selected most frequently as a numerical parameter. As the simplest process, it is possible to specify the personality type having the highest frequency of music selection as the personality of the user. For example, if the number of track selections accumulated in the statistical information reaches a certain level, for example, the personality type initially set by user input may be replaced with the personality type derived from the statistical information as described above. .

  Prior to using the car, user authentication is required. This is because, in particular, when a plurality of users are registered, different personality types are set depending on the users, and the contents of hospitality are also different. The simplest authentication method is a method in which a user ID and a personal identification number are transmitted from the portable device 1 to the vehicle side, and the hospitality decision-making unit 2 receives this and verifies the registered user ID and personal identification number. is there. In addition, biometrics authentication methods such as collation of facial photographs using a camera provided in the portable device 1, voice authentication, authentication using fingerprints, and the like can be employed. On the other hand, when the user approaches the car, only simple authentication using the user ID and the personal identification number is performed, and after unlocking, the user gets into the car and then the face camera 521, microphone 522, retina camera 526, iris described above. Biometrics authentication by the camera 527 or the vein camera 528 may be performed.

  In the present embodiment, a case where authentication is performed by collating face photographs is illustrated. As shown in FIG. 32, the face camera 521 captures each sheet 100FR, FL, RR, RL from the front, and the field of view is defined so that the upper body of each occupant is included. The seating sensor 519 is composed of a load sensor or the like embedded in the seat part of each seat 100FR, FL, RR, RL, and detects the presence or absence of the seating of the passenger. For example, when a seating sensor 519 detects a load and a face image of the occupant is detected in the field of view of the face camera 521, a method of detecting the presence of seating is employed, so that a load placed on the seat or ambient light is detected. It is possible to prevent erroneous detection due to the above. In addition, the combined use of the seating sensor 519 can ensure the accuracy of seating detection even if the face image identification accuracy is slightly lowered, which contributes to the weight reduction of the algorithm. On the other hand, as shown in FIG. 2A, the user registration unit 600 stores the face feature amount extracted from the master face image of each user in association with the user ID, and is photographed by the face camera 521 of each sheet. By comparing with the facial feature amount extracted from the facial image, it is possible to specify which user is on which seat.

  Hereinafter, a specific operation example of the hospitality system of the present invention will be described with reference to a flowchart of FIG. 13 taking a driving / staying scene as an example. First, the mental state and physical condition of the sitting user are estimated (FIG. 12: γ3, γ4). The specific method is disclosed in detail in Patent Documents 1 to 3, and only the outline will be described here. Basically, a method is employed in which a temporal change in the acquired biological condition parameter is measured, and a mental state and a physical condition are estimated from the change waveform.

  In this embodiment, as shown in FIG. 23, each mental state is based on the concept of a so-called Russell-Merabian emotional plane in which mental activity (wakefulness) is defined on the vertical axis and pleasantness is defined on the horizontal axis. Are classified and specified. Specifically, the four mental states defined by the individual quadrants of the emotion plane, namely, the swell state (high mental activity / pleasure), the healing / relaxed state (low mental activity / pleasure), the anger / excitement state ( It is defined as being divided into a plurality of stages according to the distance from the origin of the emotional plane, based on the basics of mental activity level / unpleasant) and discouraged / malacious state (mental activity level / unpleasant) The area near the origin of this emotion plane represents a neutral mental state that is not strongly biased to any of the emotional states in each quadrant, i.e., the neutral state, and as it moves away from this neutral state in each quadrant, It changes in stages to emotions that are characteristic of the quadrant, and thus to a state that is strongly controlled by desire. The individual stages of this emotion or desire are arranged according to Maslow's desire stage theory for each quadrant.

Maslow's desire stage theory is a hypothesis that human desires are divided into five stages as follows, and when lower order needs are satisfied, higher order desires are sought in order.
First stage: Physiological desires… Desires necessary for living biologically, such as appetite, sleepiness, and libido.
Phase 2: The desire for safety and stability: the desire to be protected from danger, including the future, such as clothing, housing, and money.
Stage 3: Affiliation / Affection / Social Desire ... A desire to belong to a group or be loved by family and others.
Stage 4: Desire of ego and dignity ... A desire to be recognized and respected by others.
Stage 5: Self-realization desire ... A desire for self-growth and development with the unique qualities and abilities.
However, as shown in FIG. 23, the number of stages includes “neutral state” obtained by merging the first stage and the second stage, “intermediate state” obtained by merging the third stage and the fourth stage, and “features” of the fifth stage. It is aggregated into three of the “normalized state”. Then, based on the change waveform of the temporal change measurement of the biological condition parameter described above, the mental state of each quadrant / stage is estimated.

  When “blood pressure” is adopted as the biological condition parameter, the central frequency (or peak frequency) f of the blood pressure waveform detected by the blood pressure sensor 524 and the average value A of the integrated amplitudes A1 and A2 for each section are calculated. If the frequency f is larger than the upper limit threshold fu0, it is determined that the blood pressure change being monitored is “sudden”, and if it is smaller than the lower limit threshold fL0 (> fu0), it is determined that it is “slow”. If ≧ f ≧ fL0, it is determined as “standard”. Further, the amplitude A is compared with the threshold value A0, and if A ≦ A0, it is determined that the average blood pressure level being monitored is in the “maintained” state, and if not, it is determined that it is in the “variable” state. The contribution to the determination of mental state is “slow / increase (intermediate state A ′)” → “sudden / increase (characterized state A)” in the “anger / excited state” quadrant when viewed in “slow / rapid change / direction”. "Slightly steep / decrease (intermediate state B ')" for the "disappointing / fatigue" quadrant, "Slow / decrease (characterized state B)", "Slow / increase (intermediate state C)" “)” → “Sudden / Increase (Characterized State C)” and “Healing / Relaxed State” quadrant “Slightly Steep / Decreased (Intermediate State D ′)” → “Slow / Decreased (Characterized State D)” Transition in shape. Further, regarding poor physical condition, blood pressure fluctuations are moderate.

  When “body temperature” is adopted as the biological condition parameter, a body temperature sensor 525 (infrared sensor 519) is used. Specifically, the body temperature value detected every time the sampling timing set at a fixed time interval arrives is sampled, the waveform is recorded, and a well-known fast Fourier transform process is performed to obtain a frequency spectrum, and the center of the spectrum is obtained. The frequency (or peak frequency) f is calculated. In addition, the waveform is divided into a certain number of sections, and the body temperature average value for each section is calculated. Then, for each section, the integral amplitude is calculated using the average body temperature value as the waveform center line, and the integral amplitude in each section is averaged and determined as a representative value of the waveform amplitude. It is checked whether or not the determined frequency f is larger than the upper limit threshold fu0. If it is larger, it is determined that the body temperature change being monitored is “sudden”. Further, whether or not the frequency f is smaller than the lower limit threshold fL0 (> fu0) is examined, and if it is smaller, it is determined that the body temperature change being monitored is “slow”. The contribution to the determination of mental state is “slow / increase (intermediate state A ′)” → “sudden / increase (characterized state A)” in the “anger / excited state” quadrant when viewed in “slow / rapid change / direction”. "Slightly steep / decrease (intermediate state B ')" for the "disappointing / fatigue" quadrant, "Slow / decrease (characterized state B)", "Slow / increase (intermediate state C)" “)” → “Sudden / Increase (Characterized State C)” and “Healing / Relaxed State” quadrant “Slightly Steep / Decreased (Intermediate State D ′)” → “Slow / Decreased (Characterized State D)” Transition in shape. When body temperature is stable at a higher level than normal fever (such as during cold fever) or stable at a lower level than normal fever (such as during anemia), it can also be used to identify poor physical condition.

  When “skin resistance” is adopted as the biological condition parameter, the skin resistance sensor 545 is used. Here, similarly, the skin resistance value is sampled and the waveform is recorded, and the center frequency (or peak frequency) f of the spectrum and the integrated amplitudes A1, A2,. Then, the integral amplitude A of each section is plotted against time t, and the gradient α is obtained by least squares regression. If the frequency f is larger than the upper limit threshold fu0, it is determined that the change in the skin resistance being monitored is “sudden”, and if it is smaller than the lower limit threshold fL0 (> fu0), it is determined that it is “slow”. Further, if fu0 ≧ f ≧ fL0, it is determined as “standard”. Furthermore, the absolute value of the gradient α is compared with a threshold value α0. If | α | ≦ α0, it is determined that the average skin resistance level being monitored is in a “constant” state. In the case of | α |> α0, if the sign of α is positive, it is determined that the average skin resistance level being monitored is in an “increasing” state, and if it is negative, it is determined that it is in a “decreasing” state. The contribution to the determination of mental state is "slow / decrease (intermediate state A ')" → "sudden / decrease (characterized state A)" for the "anger / excited state" quadrant when viewed in "slow / rapid change / direction" ”,“ Slightly steep / increase (intermediate state B ′) ”for the“ disappointing / fatigue ”quadrant“ → Slow / increase (characterized state B) ”, and“ slow / decrease (intermediate state C) for the “swelling state” quadrant ') "→" Sudden / Decrease (Characterized State C) "," Healing / Relaxed State "quadrant is" Slightly Steep / Increased (Intermediate State D') "→" Slow / Increased (Characterized State D) " Transition in shape. With regard to poor physical condition, mild ones are not so much reflected in the temporal change in skin resistance, but as the poor physical condition progresses, the change in skin resistance value gradually increases, so it is effective in estimating "severe poor physical condition" It is.

  When “facial expression” is adopted as the biological condition parameter, the facial camera 521 in FIG. 1 is used to capture a facial image at a predetermined sampling interval, and sequentially compare it with the master image to obtain a facial expression type (eg, “stable”). , “Anxiety / discomfort”, “excitement / anger”), and the temporal change is determined as “small change”, “increase”, “slight increase”, “rapid increase” or the like. The contribution to mental state specification is “discomfort / decrease (intermediate state A ′)” → “anger / excitement / rapid increase (characterized state)” in the “anger / excited state” quadrant when viewed in “expression type / change direction”. A) ”and“ disappointment / malaise ”quadrants for“ no expression / (small change) (intermediate state B ′) ”→“ somewhat uncomfortable / (small change) (characterized state B) ”,“ excited state ”quadrants Is "no fun / (small change) (intermediate state C ')" → "no fun / increase (characteristic state C)", "no expression / (small change) (intermediate state)" D ') "→" Slightly enjoyment / slight increase (characterized state D) ".

When “posture” is adopted as the biological condition parameter, posture signals from a plurality of seating sensors 520 embedded in the seat are used, and the center frequency (or peak frequency) f of the posture signal waveform and the integrated amplitudes A1 and A2 for each section are used. average value an of ‥ and calculates the variance sigma ^2. If the frequency f is larger than the upper limit threshold fu0, it is determined that the posture change speed being monitored is “increase”, and if it is smaller than the lower limit threshold fL0 (> fu0), it is determined that it is “decrease”. Further, the average value An of the integrated amplitude A is compared with a predetermined threshold value, and the posture movement amount is determined as “small change”, “slight increase”, or “sudden increase” (the average value An is large). The amount of posture movement tends to increase. Further, it is determined that if the value of the variance sigma ² is greater than or equal to the threshold value, attitude change is in the increasing or decreasing trend. The contribution to the identification of the god state is “increase / increase (intermediate state A ′)” → “rapid increase / increase (characterized state)” for the “anger / excited state” quadrant when viewed from “posture movement increase / decrease / speed”. A) ”,“ decreasing / stable (intermediate state B ′) ”for the“ disappointing / fatigue ”quadrant,“ sudden decrease / stable (characterized state B) ”, and“ slightly increasing / stable (intermediate state) ” “State C ′)” → “Increase / Slightly Increase (Characterized State C)” and “Healing / Relaxed State” quadrant “Decrease / Stable (Intermediate State D ′)” → “Decrease / Stable (Characterized State D)” It changes in the form. When poor physical condition occurs, gestures that occasionally change postures tend to stand out in an effort to relieve pain, and the posture movement amount tends to increase slightly. However, when the physical condition further progresses (or when extreme sleepiness is attacked), the posture becomes unstable and wobbles, and the posture movement tends to increase or decrease. At this time, the posture movement is unstable with no control of the body, so the speed of the posture movement is greatly reduced.

  When “line-of-sight angle” is adopted as the biological condition parameter, the pupil position and the face center position are specified in the above-mentioned face image, and the line-of-sight angle is calculated by calculating the blur from the front direction of the pupil with respect to the face center position. Obtain θ and obtain its time-varying waveform. Similarly, the center frequency (or peak frequency) f of the waveform and the average value An of the integrated amplitudes A1 and A2 for each section are calculated. The average value An of the integrated amplitude A is compared with a predetermined threshold value, and the amount of change in the line-of-sight angle θ is determined as “small change”, “slight increase”, or “sudden increase” (the average value An is The larger the value, the more the amount of change in the viewing angle θ tends to increase). The contribution to mental state identification is “decreasing (intermediate state A ')” → “severely decreasing (characterizing state A)”, “disappointment / malaise” in the “anger / excited state” quadrant when viewed in “Gaze movement amount” “For the quadrant,“ decrease (intermediate state B ′) ”→“ severe decrease (characterized state B) ”, for the“ rising state ”quadrant,“ stable (intermediate state C ′) ”→“ decrease (characterize state C) ” The “healing / relaxed state” quadrant changes in the form of “gradual decrease (intermediate state D ′)” → “decrease (characterized state D)”. In addition, when a poor physical condition occurs, the amount of line-of-sight movement decreases according to the degree of the defect.

When “pupil diameter” is adopted as the biological condition parameter, the iris of the user is photographed by the iris camera 527 (FIG. 1), the pupil diameter d is determined on the image, and the time-varying waveform is acquired. The pupil diameter average value of each section from the waveform dn, calculates the average value An and variance sigma ² sections separate integral amplitudes A1, A2. If the pupil diameter average value dn is larger than the threshold value d0, it is determined that the pupil is open. If not larger, it is checked whether or not the variance Σ ² is larger than the threshold Σ ²⁰ , and if larger, it is determined as “pupil diameter fluctuation”. If it is not larger, it is determined as “normal”. The pupil diameter d changes significantly according to the mental state of the user, and in particular, it is possible to estimate with high accuracy whether or not the user is in an excited state based on whether or not there is a specific pupil open state. it can. The contribution to mental state identification is “variation (intermediate state A ')” for the “anger / excited state” quadrant, “open (characterized state A)”, and “fluctuation (intermediate state)” for the “disappointment / malfunction” quadrant. B ') "→" Closed (characterized state B) "," Swelling state "quadrant" Fluctuation (intermediate state C') "→" Open (characterized state C) "," Healing / relaxed state "quadrant Changes in the form of “variation (intermediate state D ′)” → “closed (characterized state D)”.

When “steering operation state (steering)” is used as the biological state parameter, steering is sampled and evaluated during straight running. Specifically, every time the sampling timing determined at a certain time interval arrives, the current steering angle φ is read from the output of the steering angle sensor 547 (for example, φ = 0 ° in a straight traveling neutral state, either left or right) (For example, the right angle is positive and the left angle is negative.) Then, the steering angle value is acquired as a waveform, and the center frequency (or peak frequency) f and each section are obtained. , And its variance Σ ^2. If the frequency f is larger than the upper limit threshold fu0, it is determined that the change speed of the steering angle φ is “increase”, and the lower limit threshold fL0 (> fu0). ) If it is smaller, it is judged to be “decrease”, and if fu0 ≧ f ≧ fL0, it is judged to be “normal”, and if the variance Σ ² of the integrated amplitude A is larger than the threshold Σ ² 0 It is determined that the steering error is “increase”. If it is not large, it is determined as “normal.” The contribution to the mental state specification is “increase / increase (intermediate state A ′)” for the “anger / excited state” quadrant when viewed from “increase / decrease in steering error / steering speed”. "→" Increase / Increase ((characterized state A) "," Normal / normal (intermediate state B ') "→" Increase / normal (characterized state B) "for the" disappointment / malfunction "quadrant. Both the “rising state” quadrant and the “healing / relaxed state” quadrant remain normal and no noticeable tendency occurs.

  The determination (estimation) of the specific physical condition / mental state of the user is performed using the determination result relating to the temporal change of the biological condition parameter obtained in this way. In each storage device 535, as shown in FIG. 14, a plurality of specified states that are combinations of mental states and physical states to be determined by the user, and individual specified states A determination table 1601 is stored in which a plurality of user biological characteristic information acquisition means for determining that the relationship is established and a combination of temporal change states of biological state parameters that should be detected are stored in association with each other. Yes.

  Specifically, determination results (for example, “sudden decrease”, “increase”, etc.) relating to temporal changes of each biological condition parameter are read, and the individual determination results thus read are displayed on each specified state on the determination table 1601. Sequentially collate with combinations of temporal changes corresponding to. In this case, for example, for all the biological condition parameters, only the specified state in which the collated information matches the determination result may be adopted. However, when there are many biological condition parameters to be referred to, it is determined as the collated information. The result rarely matches for all the biological condition parameters, and the user's physical condition or mental condition cannot be flexibly estimated. Therefore, it is effective to consider the score (N) of the matching counter as “matching degree” and to determine the highest score, that is, the highest matching score, as the specified state.

  It should be noted that, for example, when the average blood pressure level is determined to be “variation”, the state of the same biological condition parameter is positive for establishment of a plurality of specified states (“concentration distraction” or “excited state”). In this case, the collation counter of each specified state is incremented. For example, when it is determined that the average blood pressure level is “fluctuation”, the four matching counter values N1, N4, N5, and N6 are incremented.

  On the other hand, the match / mismatch between the information to be matched and the determination result is mostly determined by comparison with the threshold value of the biological condition parameter (frequency or amplitude, etc.) as described above. / When determining the discrepancy binary (that is, whether it is white or black), the degree of deviation between the indicated value of the actual parameter and the threshold value is buried as information. However, in actuality, when matching / non-coincidence is determined by a value close to the threshold value, it is a determination of “gray”, and matching / non-coincidence is separated from the threshold value (for example, the threshold value is greatly cleared). It is inherently desirable to handle the degree of contribution to the determination result smaller than when it is determined.

  As a method of solving this, instead of adding to the verification counter only when the information to be verified and the determination result are completely matched, it is possible to approach within a predetermined range without a complete match. If the result is obtained, it is conceivable that the score is added to the verification counter while limiting the score to a lower score than in the case of perfect match. For example, when the information to be collated is “rapid increase”, 3 points are added if the determination result is “rapid increase”, 2 points are added for “increased”, and 1 point is added for “slight increase” Can be illustrated.

  Returning to FIG. 13, if the physical condition and the mental condition can be specified, the process proceeds to S <b> 2, and the physical condition reference intention parameter value corresponding to the specified physical condition is determined. In the present embodiment, the physical condition reference intention parameter is a spatial parameter composed of five components (fb1,..., Fb5) (hereinafter referred to as [fb] ≡ [fb1,. Called). Among the five components, fb1 and fb2 belong to the parameter type representing the intention strength related to the reduction of physical strength, and fb3, fb4 and fb5 belong to the parameter type representing the intention strength related to the recovery of physical strength. Specifically, fb1 corresponds to each intention of “reducing force / time” and fb2 corresponding to “reducing labor / person”. Further, fb3 corresponds to each intention of “supplement”, fb4 “rest”, and fb5 “enhance effect”.

  Next, in S3, it is a process of determining a mental system standard intention parameter value corresponding to the specified mental state. In this embodiment, the mental system standard intention parameter is a spatial parameter composed of seven components (fm1,..., Fm7) (hereinafter referred to as [fm] ≡ [fm1,. Called). Of the above seven components, fm1 and fm2 belong to the parameter type representing the intention strength relating to the unpleasant elimination, and fm3 and fm4 belong to the parameter type representing the intention strength relating to the pleasant acquisition. Further, fm5 belongs to the parameter type representing the intention strength related to feeling uplift, and fm6 and fm7 belong to the parameter type representing the intention strength related to feeling rest. More specifically, fm1 corresponds to each intention of “exclude the object (exclude the stress source)” and fm2 correspond to the intention of “defiguring (no feeling of stress)”. fm3 corresponds to each intention of “get a target (recognize a favorite thing / image)” and fm4 correspond to each intention of “make a devil (enjoy another favorite atmosphere)”. fm5 corresponds to the intention of “uplifting (the image of the car / destination is emphasized)”. fm6 corresponds to each intention of “healed” and fm7 corresponds to “increase effect”.

  The set value of each component of the physical condition reference intention vector [fb] corresponding to each physical condition state is stored in the physical condition reference intention vector setting table 601 shown in FIG. 15 (ROM of the hospitality decision making unit 2 in FIG. 1). ), And those corresponding to the determined physical condition are read and used as needed. Moreover, the mental system reference intention vector [fm] is obtained at each stage (N, A, A ′, B, B ′, C, C ′, D, D ′) of each quadrant on the emotion plane shown in FIG. It is set in a form corresponding to the mental state on a one-to-one basis, and the setting values of each component are collected in the mental system reference intention vector setting table 602 (stored in the ROM of the hospitality decision making unit 2 in FIG. 1). The one corresponding to the determined mental state is read and used at any time.

  Each component of the reference intention vector [fb] [fm] (that is, the component reference intention parameter) is set to have a larger absolute value as the user's intention to exclude the corresponding disturbance stimulus is stronger. A specific example will be described. First, regarding the physical condition reference intention vector [fb], as shown in FIG. 15, the values of the parameter types fb1 and fb2 representing the intention strength related to the reduction of physical fitness are obviously larger in the case of abnormal physical condition than in the normal condition of physical condition. (In particular, the value is the maximum value (1) when the physical condition is abnormal, and the value is slightly decreased when the physical condition is abnormal (0.8). Parameter type (fb3) indicating the intended strength related to physical strength recovery , Fb4, fb5), and the value of fb4 (resting) is set to be slightly higher during fatigue even when normal.

  Further, as shown in FIG. 16, the mental type standard intention vector [fm] has a parameter type fm5 representing the intention strength related to feeling high in the discouraged / malformed state, which is larger than the neutral state, the anger / excited state, and so on. It is set and emphasizes the intention to raise the mood. On the contrary, the parameter types fm6 and fm7 representing the intention strength regarding mood rest are set to be larger than those in the disappointment / malaise state or the neutral state in the passionate / excited state, so as to emphasize the intention to cool down the feeling of high feeling. I have to.

  It should be noted that since there is usually no user who welcomes unpleasant elements, the values of parameter types fm1 and fm2 representing the intention strength regarding unpleasantness exclusion are generally set higher. On the other hand, in emotions, excitement, discouragement, fatigue, etc., there may be no room for accepting pleasant elements due to lack of mental stability, and the parameter types fm3 and fm4 representing the intensity of intention related to enjoyment acquisition are generally set lower. .

  According to the content of the mental system reference intention vector [fm] described above, the mental system reference intention vector corresponding to the mental state biased in the uncomfortable direction in FIG. On the other hand, when it is a discouraged state, it is in a form of excitement or excitement (that is, it can be regarded as a counter-type hospitality operation). On the other hand, the mental standard intention vector corresponding to the mental state or the neutral state biased in the pleasant direction in FIG. 23 has a form in which a hospitality operation is performed to maintain the pleasant state (that is, the mental state concerned). It can be seen as a compliant hospitality operation that supports or encourages.

  Next, the process proceeds to S4 in FIG. 13 to correct the physical condition reference intention vector [fb] and the mental reference intention vector [fm] to be used. Specifically, the hospitality priority of each user entering the car is searched by the user registration unit 600 of FIG. 2A, and the physical condition correction coefficients Dfb and Dfm of the user with the highest hospitality priority are read out. The physical condition correction coefficient Dfb is for correcting the component set value of the physical condition reference intention vector [fb], and the individual difference in physical stimulation level preferred by the corresponding user (that is, prefers or weakens a stronger stimulus). Depending on the user's irritation: for example, cooling favors light and cold, or conversely sweats and favors cold.) Determined. In the present embodiment, as shown in FIG. 17, the value of the physical condition correction coefficient Dfb is determined in three stages corresponding to weak, neutral, and strong, respectively. The physical condition correction coefficient Dfb is used by multiplying each component of the vector [fb], “Neutral” is no correction (1), “Weak” is value reduction (0.8), and “Strong” is value expansion ( 1.2).

  The mental system correction coefficient Dfm is used to correct the component set value of the mental system reference intention vector [fm], and is the individual difference in desire with respect to the user's disturbance rejection speed (that is, whether it is an insidious type or an insidious type: In the case of cooling, it reflects whether the user prefers to cool slowly or prefers to cool quickly), and a specific value is determined for each user. The mental correction coefficient Dfm has a close correspondence with the personality type in the user registration unit 600, and the value of the mental correction coefficient Dfm is set to be small for a decent personality, and conversely for the impure personality. The value of the mental correction coefficient Dfm is set large. In the present embodiment, the value of the mental correction coefficient Dfb is determined in three stages corresponding to slow, neutral, and fast, respectively. The mental correction coefficient Dfb is used by multiplying each component of the vector [fm], “Neutral” is no correction (1), “Slow” is value reduction (0.8), and “Fast” is value expansion ( 1.2).

  Naturally, since the preference of the physical stimulation level is different for each user, it is difficult to find a correction condition that satisfies them uniformly. Therefore, the hospitality priority of the user (so-called “power relationship” between the users who have boarded) is introduced, and processing is performed with priority given to the correction condition of the user with the highest hospitality priority. As described above, the hospitality priority setting reflecting this “power relationship” can be input at the time of user registration, and when the priority is different between a child and an adult, the camera 521 or the like shoots. The body size of the seated user is specified by the user's image or a detection sensor attached to the seat, and hospitality is given priority according to the body size (for example, if the body size is less than the threshold, it is determined as a child) The degree may be automatically set.

  It is also possible to provide a microphone for each seat, analyze the user's conversation detected by the microphone based on known voice recognition and conversation analysis techniques, and automatically set the hospitality priority. A relatively simple method is to analyze the frequency of word appearance during conversations, for example, determine that users who use a lot of honorifics are lower in power relations, set lower hospitality priority, and conversely command or fateful tone A user with a high appearance frequency of the word to be indicated can be determined to be higher in power relationship, and the hospitality priority can be set higher. In addition, it is possible to use a lot of infant words or to judge a user who is generally speechless as a child.

  The hospitality priority may be configured to set a default value regardless of the seat position of the user and correct it according to the seating position. As shown in FIG. 30, the person sitting in the driver's seat is a person who corresponds to the host in the sense of hospitality, and the passenger seat, which is the adjacent seat, tends to be seated by a person who is relatively close to the host. In this case, the hospitality priority of the passenger seat user can be relatively increased by setting the correction amount for correcting the hospitality priority in the large direction to be larger than that of the driver seat user for the passenger seat user. As for the rear seat, it is strongly customary that the seat behind the driver's seat is considered to be the upper seat on the manners, and the amount of correction for correcting the hospitality priority in the large direction is greater for the user behind the driver seat than for the user behind the passenger seat. Can be set to In addition, the front seat and the rear seat may be somewhat separated, and the correction amount of the hospitality priority can be set smaller than the adjacent seats in consideration of the social distance. In a situation where the hospitality priority of the passenger seat user is increased, a mode in which the large correction amount on the front seat side is higher than that on the rear seat side can be exemplified.

  In addition, when a child (especially an infant or a lower grade child) sits in the rear seat, as shown in FIG. 31, there may be a case where an excitement makes noise and affects driving concentration. In this case, as an example of correspondence on the rear seat side, it is possible to set the hospitality priority to the hospitality operation (for example, video output in which the child can immerse) that seeks to calm the child in the rear seat. . On the other hand, as a correspondence example on the driver's seat side, there can be exemplified a method of increasing hospitality priority with respect to the operation of a noise canceller that suppresses noises from the rear seats and the transmission of comfortable music for calming irritation.

  FIG. 27 is a flowchart showing the flow of the hospitality priority seat specific determination process. In steps S01 to S106, the user of the driver seat, the passenger seat, and the rear seat are specified by the above-described flow, and for example, the value of the default hospitality priority registered in the above-described user registration unit 600 (FIG. 2A) is read. Then, as described above, the correction coefficient determined for each seat is multiplied by the default hospitality priority to set the hospitality priority for each seat.

  In the hospitality operation, those that can be individually handled are basically executed with different contents customized for each seat. Arbitration processing is performed. In this case, as shown in FIG. 28, for example, the intention strength of the hospitality operation described later is corrected so that the higher the hospitality priority is, the stronger the intention strength is, according to a correction coefficient determined according to the value of the hospitality priority. As a result, mediation with respect to the above-mentioned operation competition becomes possible.

  Now, each component of the corrected vectors [fb] and [fm] is developed on the intention estimation table 371 shown in FIGS. 19 and 20 as follows. The intention estimation table 371 is stored in the ROM of the hospitality decision making unit 2 and has the following structure. In other words, each scene (ride → get off → stay → drive) is arranged on the vertical axis of the matrix. For each scene, three hospitality intentions: safety (safe), convenience (easy), and comfort (comfortable) A classification item is set. Of these, five intention strength items indicated by the components of the physical condition standard intention vector (parameter) are assigned to the convenience (easy) hospitality intention classification items (FIG. 20), and the comfort (comfort) hospitality intention Seven intention strength items indicated by each component of the mental system reference intention vector (parameter) are assigned to the classification items (FIG. 19).

  On the other hand, on the horizontal axis of the matrix, disturbance stimulus items divided into at least three attributes of the tactile, visual, and auditory systems (in addition, there is an olfactory sense, but it is out of the figure in the figure) are arranged. doing. The disturbance stimulus includes an in-vehicle disturbance and an in-vehicle disturbance. Here, the case of the in-vehicle disturbance is illustrated. The tactile disturbance stimulus includes airflow temperature, vehicle interior temperature, humidity, and pressure (vibration). The visual system disturbance stimulus includes light (illuminance). Furthermore, there are sounds in auditory disturbance stimuli. In either case, the value is acquired from the input of the corresponding sensor (FIG. 13: S5).

  Each type of disturbance stimulus value is detected in both positive and negative directions with respect to the neutral value (specifically, each disturbance stimulus value is set in the positive direction with the neutral value set to zero. Detected in 5 steps of +0.5, +1.0 and -0.5, -1.0 in the negative direction). All of them change in a way that goes back and forth between mutually contradictory states, such as bright and dark, hot and cold, and noisy and quiet. Specifically, with regard to airflow temperature, its intensity (corresponding to amplitude: degree) changes between hot → (warm, neutral, cool) → cold, and (the temperature change obtained as a result of functional operation) ) Velocity (corresponding to frequency) varies between fast → (fast, neutral, slow) → slow. As for the temperature inside the vehicle, its strength changes between hot → (neutral) → cold. Humidity changes between wetness → (wet, neutral) → dry. As for vibration, the intensity changes between strong → (neutral) → weak, and the speed (frequency) changes between high → (neutral) → low. For light (illuminance), the intensity changes between dazzling → (neutral, dim) → dark, and the speed (frequency (color of light)) changes between warm → (neutral) → cold. To do. As for the sound, its intensity changes between Mayan → (neutral) → quiet, and the speed (frequency) changes between lively → (neutral) → calm.

  As described above, each component (reference component intention parameter) of the reference intention vectors [fb] and [fm] is set to have a larger absolute value as the user's intention to exclude the corresponding disturbance stimulus is stronger. Is calculated as a reference intention parameter on each cell of the intention estimation table 371 with a corresponding disturbance stimulus value (FIG. 13: S6). If the corresponding disturbance stimulus value is a neutral value (that is, zero), the product value (intention strength parameter) is also zero (the blank cell in FIGS. 19 and 20 indicates the value of the intention strength parameter). Is zero). This means that the user's intention to exclude the corresponding disturbance stimulus does not exist (that is, the disturbance stimulus state most comfortable for the user is realized without doing anything), and the disturbance stimulus is suppressed. The intention to suppress (or prohibit) the operation of the corresponding function is reflected.

  If the absolute value of the corresponding disturbance stimulus value is not zero, the intention intensity parameter is not zero, and it is interpreted that the larger the absolute value is, the larger the intention to exclude the disturbance stimulus is. However, the intention intensity parameter can be defined by the reciprocal of the product of each component of the reference intention vectors [fb] and [fm] and the corresponding disturbance stimulus value, or can be defined by the ratio. Needless to say, the relationship between the value and the magnitude of the disturbance stimulus exclusion intention varies depending on the calculation definition.

  The sign of the intention strength parameter matches the sign of the disturbance stimulus, and the sign is associated with the direction of control of the corresponding function. That is, the disturbance stimulus value changes between the two states that are opposite to each other as described above, and between the two states, a neutral state that is neither of them (that is, it is just right that is neither too bright nor too dark). Brightness condition, comfortable temperature condition that is neither too hot nor too cold, or normal acoustic condition that is neither too loud nor too quiet), and the disturbance stimulus is biased to either of the above two conditions based on the neutral condition. Is uniquely represented by the sign.

  As described above, the reference intention parameter indicates, in absolute value, how much the user has a desire to exclude a disturbance stimulus deviating from the neutral value, and is used to calculate the product of the disturbance stimulus value. The direction in which the disturbance stimulus is canceled (cancelled) is controlled by the sign of the intention strength parameter calculated by. That is, it is dark when bright, bright when dark, cool when hot, and warm when cold. Also, if you are crammed in May, you should be quiet or control the direction to deceive the May ugliness.

  Specific function selection / determination is performed using the principle search table 372 shown in FIG. 21 and the function extraction table 373 shown in FIG. 22 (FIG. 13: S7, S8). The principle search table 372 has the same cell structure as the intention estimation table 371, but each cell stores principle specifying information for specifying the principle of the function as function specifying information. The function extraction table 373 is a two-dimensional table associated with the principle search table 372 in which each principle specifying information and individual functions are associated with each other. Function adoption priority information indicating the degree is stored.

  Hereinafter, a more specific example will be described by using a part related to the mental system standard intention vector as a representative (the basic processing flow is the same for the part related to the physical system standard intention vector). FIG. 19 shows the setting part of the mental system reference intention vector [fm] extracted from the intention estimation table 371, and FIG. 21 shows the corresponding part extracted from the principle search table 372. In FIG. 19, the set values of the components [fm1, fm2, fm3, fm4, fm5, fm6, fm7] of the mental system intention vector [fm] are [1, 1, 0.2, 0.2, 0,. 2, 0.8, 0.8], and [0.8, 0.8, 0.16, 0.16, 0.16, 0.64,. 64] is set on the intention estimation table 371. On the other hand, for the disturbance, for example, regarding the intensity of light, which is a visual disturbance, a minimum value of -1.0 in five steps is detected, so the setting value of the intended intensity parameter of the corresponding cell is [-0.8. , -0.8, -0.16, -0.16, -0.16, -0.64, -0.64].

  If the corresponding part of the principle search table 372 in FIG. 21 is seen, as the principle specifying information of the visual system, for example, in the intention item of unpleasant elimination → stress source elimination → communication enhancement, “internal display” (reason system) and “ "Inside illumination" (sensitivity system) is stored (cells that are blank on the principle search table 372 indicate that there is no corresponding functional principle). The former is associated with a function for turning on a display of a meter, an indicator, a car navigation device, a car stereo, or the like at night (or dark place) on the function extraction table 373 in FIG. In FIG. 22, the display is omitted). Further, the latter is vehicle interior illumination. On the function extraction table 373 in FIG. 22, in the column of in-vehicle action → visual action → illumination, up / down (ie, foot and ceiling) in the front and rear, left and right seats. ) For the illuminations arranged to illuminate each position (FIG. 1: in-vehicle illumination 511), it is specified to select the corresponding function with the priority “3”. The setting value of the corresponding intention strength parameter is −0.8 (pretty dark), and the hospitality execution control unit drives and controls the display and illumination brightly with the light emission output corresponding to +0.8 in order to cancel this. (FIG. 13: S9).

  In FIG. 19, regarding the air flow temperature (intensity) of the tactile disturbance, five levels of small and medium values −0.5 are detected. Therefore, the setting value of the intended intensity parameter of the corresponding cell is [−0.4, −0.4, −0.08, −0.08, −0.08, −0.32, −0.32].

  When the corresponding part of the principle search table 372 in FIG. 21 is seen, as the principle specifying information of the tactile system, for example, “inner air conditioner” (sensitivity system) is included in the intention item of unpleasant elimination → stress source elimination → thickness / coldness elimination. Stored. This is linked to the air conditioner function on the function extraction table 373 in FIG. 22 (however, the display is omitted in FIG. 22). The setting value of the corresponding intention strength parameter is −0.4 (cool), and the hospitality execution control unit controls the air conditioner so that the vehicle interior is warmed by the heating output corresponding to +0.4 in order to offset this. Perform (FIG. 13: S9).

  In addition, “inner lighting” (sensitivity system) is stored as the principle specifying information in the intention item of eliminating unpleasantness → defiguring → no feeling of stress. The corresponding function is adjustment of the illumination color, and the illumination color is changed so as to shift to the long wavelength side (yellow, amber, red, pink, or white light with these colors). This is because these illumination colors are warm colors and are excellent in the effect of warmth (or contribute to the uplifting of the mood).

  Furthermore, “Principal display” (reason system) is stored as principle specifying information in the intention item of “Pleasant acquisition → Preferable things / images available → Provide hobby information”. On the function extraction table 373 in FIG. 22, for example, an information output function of the car navigation device (for example, guidance information of a facility (such as a coffee shop or a pool) that provides a cool feeling) is linked (however, the display is omitted in FIG. 22). ) “Internal sound generation” (Kansei system) is stored as the principle identification information in the intention item for sending out the ambience layer. On the function extraction table 373 in FIG. 22, for example, a sound effect output function that invites a cool feeling such as murmur, wave, and wind is linked (however, the display is omitted in FIG. 22).

  In FIG. 19, regarding the sound (intensity) of the auditory system disturbance, five levels of local maximum value +1.0 (that is, the equivalent of May) is detected, so the setting value of the intention intensity parameter of the corresponding cell is [0. .8, 0.8, 0.16, 0.16, 0.16, 0.64, 0.64]. If the corresponding part of the principle search table 372 in FIG. 21 is seen, “internal sound prevention” (Kansei system) is stored as the principle identification information of the auditory system, for example, in the intention item of unpleasant elimination → stress source elimination → noise elimination. ing. This is linked to the noise canceller 1001B on the function extraction table 373 in FIG. 22 (however, the display is omitted in FIG. 22). The corresponding setting value of the intention strength parameter is 0.8, and the hospitality execution control unit controls the noise canceller 1001B so that the vehicle interior noise is canceled with an output corresponding to +0.8 in order to cancel this. (FIG. 13: S9).

  In addition, “internal music generation” (sensibility system) is stored as the principle specifying information in the intention item “Pleasant acquisition → You can get what you want / images → Provide hobby information”. On the function extraction table 373 in FIG. 22, the music output function of the car stereo is linked (however, the display is omitted in FIG. 22). The music source performs appropriate music selection according to physical condition and mental state each time (FIG. 13: S9).

  Next, in the ROM of the hospitality decision making unit 2 in FIG. 1, there is a state transition pattern for guiding the user mental state according to a certain scenario, including the case where the user mental state is relatively stable in the neutral state. Multiple patterns are stored. Each state transition pattern is determined so as to reach the end point mental state from the base mental state through one or more progress mental states, and a series of guidance hospitality operations respectively adapted to transition / guidance between adjacent mental states. It is stored in association with the contents. Then, according to the above-described method, the current mental state of the user is detected, and a state transition pattern having a base mental state that matches the detected current mental state is selected on the hospitality decision making unit 2 side. The hospitality decision-making unit 2 sequentially reads out the guidance hospitality operation contents according to the selected state transition pattern, and guides the hospitality operation for causing the hospitality execution control unit 3 to sequentially perform the guidance hospitality operation. Make a command. In the hospitality operation unit 3 side, the direction of mental induction is determined spontaneously by the system side based on the state transition pattern selected according to the detected current mental state. A series of hospitality operations related to guidance can be actively received from the system side in an appropriate flow according to the situation.

  FIG. 23 shows a specific example of the state transition pattern. As described above, the mental state of the user is higher than the neutral state (N), the anger / excited state (A: the second quadrant), and the neutral state (N), which are higher in mental activity than the neutral state (N). From the discouraged state (B: 3rd quadrant) where the mental activity is low in the uncomfortable direction, the rising state (C: 1st quadrant) where the mental activity is higher in the pleasant direction than the neutral state (N), and from the neutral state (N) Is also healed (D: 4th quadrant) with low mental activity in a pleasant direction. Also, in each quadrant, depending on the distance from the origin of the emotion plane, the neutral state (N) → the intermediate state (A ′, B ′, C ′, D ′) → the characterization state (A, B, C, D) is divided into three stages.

The stored state transition pattern includes the following types.
(1) Corrective dedicated state transition pattern The anger / excited state (A) is set as an unpleasant base mental state, the intermediate state (A ′) (intermediate corrected mental state) is set as a transitional mental state, and then the neutral state (N) is reached. Guide (SN _12R ). Also, the discouraged / failed state (B) is set as the unpleasant base mental state, the intermediate state (B ′) (intermediate corrected mental state) is set as the elapsed mental state, and then the neutral state (N) is led (SN _22R ). It becomes a mental induction scenario that raises a depressed mental state.

(2) Regression type state transition pattern There are two types, a swell type transition pattern and a healing type transition pattern. The swell type transition pattern uses the neutral state (N) as a base mental state, and returns to the neutral state (N) through the progress mental state consisting of the swell state (C) (SN _11F + SN _11R ). The pattern has a neutral state (N) as a basic mental state, a first pattern SN _11F that leads to a characteristic state (C) through an intermediate state (C ′), and a characteristic state (C) as a basic mental state. The second pattern SN _11R returns to the neutral state (N) through the intermediate state (C ′). On the other hand, in the healing type transition pattern, the neutral state (N) is set as the base mental state, and the state is returned to the neutral state (N) through the progress mental state composed of the healed state (D) (SN _21F + SN _21R ). The pattern has a neutral state (N) as a basic mental state, a third pattern SN _21F that leads to a characteristic state (D) through an intermediate state (D ′), and a characteristic state (D) as a basic mental state. The fourth pattern SN _21R returns to the neutral state (N) through the intermediate state (D ′).

  In any type of state transition pattern, the intermediate state is passed through the elapsed mental state (A ′, B ′) so that the user can comfortably move to the end mental state without feeling unnatural and sudden mental state change. Can guide smoothly.

  The details of each state transition pattern are shown in FIGS. 24 to 26 in association with setting examples of the hospitality operation. In the hospitality operation corresponding to each mental state, as described above, the mental standard intention vector [fm] corresponding to the specified mental state is read out, the component is corrected as necessary, and the intention estimation table 371 ( 19 and 20), and after that, the function used for the final hospitality operation is selected through the principle search table 372 (FIG. 21) and the function extraction table 373 (FIG. 22). In order to switch the hospitality operation corresponding to the current mental state to the hospitality operation corresponding to the target mental state, the mental system reference intention vector [fm] corresponding to the current mental state is set, and then the mental system reference is performed. This is done by switching the intention vector [fm] to that corresponding to the target mental state. In this case, the contents of the intention estimation table 371, the principle search table 372, and the function extraction table 373 need to be determined as appropriate so that the desired hospitality operation can be obtained by switching the mental reference intention vector [fm]. However, it is very complicated to strictly explain the change in the hospitality operation according to the state transition pattern according to the algorithm using the tables 371, 372, and 373, and it is difficult to roughly grasp the change in the hospitality operation. In the following description, the state transition pattern and the specific hospitality operation (function) content finally specified according to the above algorithm will be described in a directly associated form.

The upper column of FIG. 24 shows an example of the modified dedicated type state transition pattern SN _12R in which the second quadrant of FIG. 23, that is, the anger / excited state (A) is the unpleasant base mental state. In the anger / excited state (A) which is the basic mental state, a compliant hospitality operation mainly for supporting (that is, accepting) the anger / excited state is incorporated as a preparation hospitality operation. Specifically, examples include generation of a scent that invites excitement (rosemary, sweet orange, etc.) and air conditioning operation (such as slightly increasing the air conditioning temperature and slightly increasing the humidity). In addition, the output of information that excludes the tastes of others is also effective in supporting the anger / excited state of the target user. For example, in the case of sound information output, an operation of selecting and outputting music that is a preference of the target user and that does not belong to the preference of the passenger on the ride at all can be exemplified.

  In order to guide this anger / excited state (A) to an intermediate state (A ′) in which the anger / excited state is somewhat calmed down, a first guidance hospitality operation (A → A ′) is performed. Specifically, the output of information that excludes the taste of others (for example, music: video and hobby information may be output) is slightly relieved from the exclusion of preferences of others (for example, passengers who ride with passengers are not disliked) Alternatively, the information is switched to the target user's favorite information and output (preference-suitable hospitality operation). In addition, the scent generation, air conditioning, and the like are switched from a compliant hospitality operation to a counter-type hospitality operation (a calming fragrance (such as neroli, roman chamomile, lavender, and jasmine), temperature and humidity that does not cause discomfort).

  After that, the second guidance hospitality operation (A ′ → N) for shifting from the intermediate state (A ′) to the neutral state (N) is performed. In this case, the preference suitable type hospitality operation is incorporated following the first guidance hospitality operation (A → A ′), and the incorporation of the counter-type hospitality operation is more limited than the first guidance hospitality operation. Specifically, information suitable for the target user's specific taste is suppressed (volume reduction or stop in music, etc.), fragrance generation that performs counter-type hospitality is stopped, and ventilation by air conditioning (for example, to discharge residual aroma) Outside air intake mode) or deodorizing operation is executed. In addition, a video (or music) that can be enjoyed in common by all users including the target user is output as a common taste-preferred type hospitality operation. It is also effective to switch the illumination color to a bright color (such as greenish white) that promotes calming.

  If it is possible to shift to the neutral state by the above guidance, it is important to ensure that the hospitality operation adapted to the neutral state is not excessively stimulated by a stable mental state. It is a hospitality operation that concentrates on quietness, visibility, etc.) and elimination of unpleasant factors (such as elimination of heat, cold, and humid conditions).

Next, the lower column of FIG. 24 shows an example of the modified dedicated state transition pattern SN _22R in which the third quadrant of FIG. 23, that is, the discouraged / failed state (B) is the unpleasant base mental state. In the disappointment / fatigue state (B), which is a basic mental state, a compliant hospitality operation for mainly supporting (ie, accepting) a depressed state is incorporated as a preparation hospitality operation. Specific examples include scent generation operations (such as sweet marjoram, lavender sal, lavender, etc.) to be calmed down, and air conditioning operations (such as slightly lowering the air conditioning temperature and slightly lowering the humidity). In addition, slightly excluding other people's preferences (for example, the degree that passengers feel “not disliked” or “not bad”), and outputs the target user's preference information (music, video, hobby information, etc.) together I am doing so.

  A first guidance hospitality operation (B → B ′) is performed to induce this discouraged / malored state (B) to a slightly activated intermediate state (B ′). Specifically, a video that can be enjoyed in common by the previous user including the target user is output as a common taste-preferred type hospitality operation. Also, the music outputs light and crisp music (such as pop music) that raises the depressed mood, and the scent generation and air conditioning, etc., stimulate the spiritual aroma (rosemary, near uri, thyme, grapefruit, etc.) Switch to a counter-type hospitality operation, such as a slightly higher temperature setting.

  Then, after that, the second guidance hospitality operation (B ′ → N) for shifting from the intermediate state (B ′) to the neutral state (N) is performed, the content of which is the processing in the second quadrant of FIG. Fits almost the same. Here, following the first guidance hospitality operation (B → B ′), a video (or music) that can be enjoyed in common by all users including the target user is output as a common preference-optimized hospitality operation. .

FIG. 26 shows an example of the healing type transition pattern SN _21F + SN _21FR in the fourth quadrant of FIG. 23, that is, among the regression type state transition patterns having the neutral state (A) as the base mental state. In the first pattern SN _21F , when the user is guided from the neutral state (N) to the intermediate state (D ′), the target user has maintained a cooperative mental state with other users due to the occurrence of self-preference. A separation hospitality operation is performed to separate from the surroundings. In this case, the common taste-preferred type hospitality operation (music, video, etc.) is gradually changed to a taste-preferred type hospitality operation (music, video, etc.) specific to the target user. In addition, the generation of common taste fragrance is stopped and ventilation is performed. Then, when guiding from the intermediate state (D ′) to the characterization state (D), it is set as a process in which the ratio of the taste suitable type hospitality operation (music, video, etc.) specific to the target user is further increased. Then, when the characterization state (D) is reached, while continuing the taste-preferred type hospitality operation, the output operation of scents (clarisage, etc.) convenient for healing self-immersion, temperature (slightly lower) and humidity ( Perform a slightly fine adjustment).

In the second pattern SN _21R , the operation in the characterization state (D) is the same as that in the first pattern SN _11F . In order to return the characterized state (D), which has progressed in the self-immersion state, to the intermediate state (D ′) again, a kind of awakening action is required, and a strong scent (Rosemary, and further, for awareness) Outputs such as slight ammonia odor are effective. Further, when returning from the intermediate state (D ′) to the neutral state (N), the taste preference type hospitality operation (music, video, etc.) specific to the target user is gradually reduced to switch to the common preference type hospitality operation, and thereafter Then, the operation proceeds to the hospitality operation corresponding to the above-described neutral state (N).

FIG. 26 shows an example of the swell type transition pattern SN _11F + SN _11FR in the first quadrant of FIG. 23, that is, the regression type state transition pattern having the neutral state (A) as the base mental state. In the third pattern SN _11F , when the user is guided from the neutral state (N) to the intermediate state (C ′), the emphasized state is maintained while the target user has maintained a cooperative mental state with other users. As a process of raising the height, a common taste-preferred type hospitality operation (music, video, etc.) is introduced, and a bright color (yellowish white, etc.) that promotes activation is switched. Also, light and crisp music that contributes to excitement (for example, pop music) is output, and the temperature setting is made slightly higher. And when guiding from the intermediate state (C ′) to the characterization state (C), when the target user becomes a speaker, a process of cutting ambient noise with a noise canceller so that the speaker can easily hit a spot. Do. When the characterizing state (C) is reached, the conversation of surrounding users is analyzed, a word suggesting the target user's preference information is extracted in the conversation, and hobby information reflecting the contents of the word, A video or music is output, and a hospitality operation is performed to make the target user stand out (or lead to self-revelation) by extracting topics from a circle of many people.

In the fourth pattern SN _11R , the operation in the characterization state (C) is the same as that in the third pattern SN _11F . To return to the intermediate state (C ') again from the characterization state (C), which has been developed to include self-exhibition involving other people, the system side again concentrates on the target user in order to decouple other users. It is effective to sequentially switch to the operation, and finally, a process of transitioning to preference information (music, sound effects or fragrance generation) unique to the target user is performed. In addition, when returning from the intermediate state (C ′) to the neutral state (N), the taste preference type hospitality operation (music, video, etc.) unique to the target user is gradually reduced to switch to the common preference type hospitality operation, and thereafter Then, the operation proceeds to the hospitality operation corresponding to the above-described neutral state (N).

FIG. 29 shows a selection algorithm for a state transition pattern according to the current mental state. In S201, the current mental state of the target user is specified according to the method described above. If it is neutral (N) in S202, the process proceeds to S211. Basically, the process follows a regression type state transition pattern in which the uplifting type transition pattern SN _11F + SN _11FR and the healing type transition pattern SN _21F + SN _21FR are alternately repeated. By default, the swell type transition pattern SN _11F + SN _11FR is determined to be selected, and in S211, if the previous type is healed and the type transition pattern SN _21F + SN _21FR is _selected , the process proceeds to S212 because the swell condition is satisfied, and the first pattern Select SN _11F . On the other hand, if the previous time is the swell type transition pattern SN _11F + SN _11FR , it is determined that the swell condition is not satisfied, the process proceeds to S213, and the second pattern SN _11R is selected.

If the swell (N) in S203 means that the first pattern SN _11F is completed, the process proceeds to S210, and the second pattern SN _11R is selected. In the case of healing (D) in S204, it means the completion of the third pattern SN _21F , so the process proceeds to S210, and the fourth pattern SN _21R is selected. If the excitement (A) is S205, the process proceeds to S207, and the correction dedicated state transition pattern SN _11R is selected. If the disappointment (B) is S206, the process proceeds to S208, and the correction dedicated state transition pattern SN _22R is selected. . If none of the states is applicable, it means that each of the intermediate states (A ′, B ′, C ′, D ′) is satisfied. To do. The above process is repeatedly executed at regular intervals while monitoring the current mental state. In this case, if the modified dedicated state transition pattern SN _11R is executed, the neutral state (N) is finally derived. Thereafter, the neutral state (N) is used as a starting point to obtain a regression state transition pattern. It will be a process to migrate. This aspect can be similarly applied to the driver and the passenger. On the other hand, the transition to the regression type state transition pattern is performed only for the driver, and for the passenger, the regression type state transition pattern is discontinued by the first patterns SN _11F to SN _21F , and the correction dedicated state transition is performed. You may make it implement as a pattern.

  The hospitality action for guiding to each mental state (N, A ′, A, B ′, B, C ′, C, D ′, D) is performed by monitoring the current mental state, It can be implemented in a feedback control mode in which it is confirmed whether or not the target mental state is reached, and if not reached, the hospitality operation is continued. In this case, in order to stabilize the reached mental state, the hospitality operation corresponding to the target mental state is continued for a certain time (for example, 10 minutes to 1 hour is appropriate), and then the next mental state is reached. It is desirable to switch to guided hospitality operation. In the above-mentioned regression type state transition pattern, the duration of the hospitality operation for maintaining the raised state when the raised state (C) is reached is the healing time when the healing state (D) is reached. It is set to be shorter than the duration of the hospitality operation for maintaining the state (for example, about 1/2 to 2/3 of the former).

The block diagram which shows an example of an electrical structure of the user service system for motor vehicles of this invention. Explanatory drawing which shows an example of a user registration part. The block diagram which shows an example of the electrical structure of vehicle interior lighting. The conceptual diagram which shows the structural example of the lighting control data of an illuminating device. The circuit diagram which shows an example of the illuminating device using a light emitting diode. The figure which shows the mixing ratio of each illumination light, and the relationship of luminescent color of RGB full color illumination. The block diagram which shows an example of the electrical constitution of a car audio system. The conceptual block diagram which shows the structural example of a noise canceller. The block diagram which similarly shows an example of a hardware configuration. FIG. 3 is a circuit diagram showing an example of hardware for generating a posture signal waveform. The image figure of various specified states. The conceptual diagram which shows the content of a music source database. The flowchart which shows the whole flow of hospitality processing. The flowchart which shows an example of a function selection process. The figure which shows an example of the determination table. The figure which shows an example of a physical condition type | system | group reference | standard intention vector setting table. The figure which shows an example of a mental system standard | designated intention vector setting table. The figure which shows the example of a setting of a physical condition system correction coefficient. The figure which shows the example of a setting of a mental system correction coefficient. The figure which shows an example of the intention estimation table. The figure which shows another part of an intention estimation table. The figure which shows an example of a principle effect | action table. The figure which shows an example of a function extraction table. The figure which maps and shows the example of a setting of a state transition pattern on an emotion plane. The 1st figure which shows the specific example of a state transition pattern. Similarly second figure. 3rd figure. The flowchart which shows the hospitality priority setting algorithm. The figure which illustrates the relationship between hospitality priority and an intention intensity | strength correction coefficient. The flowchart which shows the selection algorithm of a state transition pattern. The figure which shows the correction | amendment concept according to the seat of the hospitality priority. The schematic diagram which shows the example of hospitality operation | movement when a child seats in a backseat. The schematic diagram which shows the example of vehicle interior arrangement | positioning of a face camera and a seating sensor.

Explanation of symbols

2 Hospitality decision-making unit (mental state identification means, state transition pattern storage means, state transition pattern selection means, guidance hospitality operation command means)
3 Hospitality Execution Control Unit 100 Automotive User Hospitality System 502-517,534 Hospitality Operation Unit 511 Car Interior Lighting (Lighting Device)
521 face camera (user biometric information acquisition means)
523 Pressure-sensitive sensor (User biological characteristic information acquisition means)
519 Infrared sensor (user biometric information acquisition means)
525 Body temperature sensor (user biometric information acquisition means)
524 Pulse sensor (User biometric information acquisition means)

Claims (14)

A plurality of hospitality operation units for performing a hospitality operation to assist a user in using a car or to entertain a user;
User biometric information acquisition means for acquiring user biometric information;
Mental state specifying means for specifying the mental state of the user based on the content of the acquired user biometric information,
A hospitality decision-making unit that determines the content of the hospitality operation by the hospitality operation unit based on the identified mental state;
A hospitality execution control unit that controls the operation of the corresponding hospitality operation unit so that the hospitality operation is performed according to the determination content of the hospitality decision making unit,
The hospitality decision-making unit
A series of predetermined state transition patterns for inducing the mental state of the user from the base mental state to the end mental state through one or more progress mental states respectively adapted to the transition / guidance between adjacent mental states State transition pattern storage means for storing a plurality of patterns in association with the content of the guidance hospitality operation;
A state transition pattern selecting means for selecting a state transition pattern having a base mental state that matches the current mental state of the user specified by the mental state specifying means with the state transition pattern storage means;
Guidance that sequentially reads out the contents of the guidance hospitality operation according to the selected state transition pattern and instructs the hospitality execution control unit to cause the corresponding hospitality operation unit to sequentially perform the guidance hospitality operation. Hospitality operation command means;
Equipped with a,
The mental state specifying means has a predetermined neutral state, an anger / excited state in which the mental activity is higher in an uncomfortable direction than the neutral state, and a mental activity level higher than that in the neutral state. Distinguishing and discriminating from disappointing low disappointment,
The state transition pattern storage means is a state transition pattern in which the anger / excited state or the discouraged state is a base mental state, and an intermediate calmed state in which the uncomfortable side base mental state is corrected in a pleasant direction is passed as the elapsed mental state. After that, memorize the modified dedicated state transition pattern leading to the neutral state,
The modified dedicated state transition pattern has a first guiding hospitality operation for transitioning from the unpleasant base mental state to the intermediate calming state, and calms down if the base mental state is the anger / excitement state When the base mental state is the discouraged state, a counter-type hospitality operation that excites it is incorporated into the first guidance hospitality operation,
The modified dedicated state transition pattern includes a preparatory hospitality operation that precedes the first guidance hospitality operation in which the counter-type hospitality operation is incorporated, and a compliant hospitality operation that supports or promotes the base mental state is the preparation. automotive user hospitality system characterized Rukoto a built in hospitality operation. The counter-type hospitality operation, to calm down the anger-excitement, or contributing aroma generating operation to activate the discouraged state, automobile users hospitality of claim 1 comprising a air-conditioning operation and the sound information output system. The base mental condition the user specific automotive user hospitality system according to specific taste adaptive hospitality operation is being further incorporated in the first induction hospitality operation claim 1 or 2 adapted to the preference regardless of the type of. The modified dedicated state transition pattern has a second guidance hospitality operation for transitioning from the intermediate calm state to the neutral state, and the second guidance hospitality operation is performed after the first guidance hospitality operation, and the specific preference is matched. The vehicle user hospitality system according to claim 3 , wherein a mold hospitality operation is incorporated, and the incorporation of the counter-type hospitality operation is more limited than the first guidance hospitality operation. The modified dedicated state transition pattern has a second guidance hospitality operation for transitioning from the intermediate calm state to the neutral state, and the second guidance hospitality operation is specified on the assumption that a plurality of users are present. Incorporation of the counter-type hospitality operation targeting the user is more limited than the first guidance hospitality operation, and the common preference preferred type hospitality operation that matches the common preference of the plurality of users is more than the first guidance hospitality operation. The automobile user hospitality system according to claim 1 or 2 , which is incorporated in an facilitated form. The modified dedicated state transition pattern has a first guidance hospitality operation for transitioning from the uncomfortable side base mental state to the intermediate calmed state, and the first guidance hospitality operation depends on the type of the base mental state. A unique preference suitable type hospitality operation adapted to the user's specific preference is incorporated, and the second guidance hospitality operation is limited in incorporation of the unique preference preferred type hospitality operation than the first guidance hospitality operation. The automobile user hospitality system according to claim 5 . 5. The automobile user hospitality system according to claim 3, wherein the unique taste-preferred type hospitality operation is an output operation of music or visual information adapted to the user's preference. The mental state specifying means includes a predetermined neutral state as a mental state of the user, a rising state having a higher mental activity in a pleasant direction than the neutral state, and a mental activity in a more pleasant direction than the neutral state. Distinguishing the low healing state from each other,
The state transition pattern storage means stores an improved dedicated state transition pattern that leads to the swell state or the healed state with the neutral state as a base mental state,
The state transition pattern selection means, automotive user hospitality system according to claim 3 or claim 4 Following the selection of the modified devoted type state transition pattern formed by the selectable said improved focus type state transition pattern. When multiple users are present, the current mental state is specified for each user and the corresponding state transition pattern is selected, and the guidance hospitality operation is individually performed for each user according to each state transition pattern. And
9. The vehicle according to claim 8 , wherein the process of selecting the improved dedicated state transition pattern following the selection of the modified dedicated state transition pattern is performed only for a user of a seat excluding the driver's seat. User hospitality system. The mental state specifying means includes a predetermined neutral state as a mental state of the user, a rising state having a higher mental activity in a pleasant direction than the neutral state, and a mental activity in a more pleasant direction than the neutral state. Distinguishing the low healing state from each other,
The state transition pattern storage means, as the state transition pattern having the neutral state as a base mental state, returns to the end point mental state including the neutral state through the progress mental state including the swelled state or the healed state. The automobile user hospitality system according to any one of claims 1 to 9, wherein a regression type state transition pattern is stored. The induction hospitality operation command means, relative to the hospitality control unit of claim 1 0, wherein performing induction hospitality operation command for the contents of the induction hospitality operation according regression type state transition pattern periodically repeated Hospitality system for automobile users. The regression type state transition pattern is set such that one cycle when the progress mental state is set to the healed state is set longer than one cycle when the progress mental state is set to the swell state. The user service system for automobiles according to 1 . The regressive state transition pattern is a guidance hospitality operation for transitioning from the neutral state to the healed state, a scent generation operation that contributes to improving the user's self-concentration state, an air conditioning operation, a light output operation, and claim 1 1 or automotive user hospitality system of claim 1 2 further comprising a self-focused promoting hospitality operation at least one of sound information output is incorporated. The regression type state transition pattern is a common-preferred type hospitality operation that matches the common preference of a plurality of users when a plurality of users are present as a guidance-type hospitality operation for transitioning from the neutral state to the swell state. automotive user hospitality system according to claim 1 2 or claims 1 to 3 is incorporated.

Images